Author: Aaron Davidson <aaron@databricks.com>
Closes#3155 from aarondav/conf and squashes the following commits:
7045e77 [Aaron Davidson] Add mesos comment
4770f6e [Aaron Davidson] [SPARK-4370] [Core] Limit number of Netty cores based on executor size
The wholeFile() can not read compressed files, it should be, just like textFile().
Author: Davies Liu <davies@databricks.com>
Closes#3005 from davies/whole and squashes the following commits:
a43fcfb [Davies Liu] remove semicolon
c83571a [Davies Liu] remove = if return type is Unit
83c844f [Davies Liu] Merge branch 'master' of github.com:apache/spark into whole
22e8b3e [Davies Liu] support compressed file in wholeTextFile
Let's give this another go using a version of Hive that shades its JLine dependency.
Author: Prashant Sharma <prashant.s@imaginea.com>
Author: Patrick Wendell <pwendell@gmail.com>
Closes#3159 from pwendell/scala-2.11-prashant and squashes the following commits:
e93aa3e [Patrick Wendell] Restoring -Phive-thriftserver profile and cleaning up build script.
f65d17d [Patrick Wendell] Fixing build issue due to merge conflict
a8c41eb [Patrick Wendell] Reverting dev/run-tests back to master state.
7a6eb18 [Patrick Wendell] Merge remote-tracking branch 'apache/master' into scala-2.11-prashant
583aa07 [Prashant Sharma] REVERT ME: removed hive thirftserver
3680e58 [Prashant Sharma] Revert "REVERT ME: Temporarily removing some Cli tests."
935fb47 [Prashant Sharma] Revert "Fixed by disabling a few tests temporarily."
925e90f [Prashant Sharma] Fixed by disabling a few tests temporarily.
2fffed3 [Prashant Sharma] Exclude groovy from sbt build, and also provide a way for such instances in future.
8bd4e40 [Prashant Sharma] Switched to gmaven plus, it fixes random failures observer with its predecessor gmaven.
5272ce5 [Prashant Sharma] SPARK_SCALA_VERSION related bugs.
2121071 [Patrick Wendell] Migrating version detection to PySpark
b1ed44d [Patrick Wendell] REVERT ME: Temporarily removing some Cli tests.
1743a73 [Patrick Wendell] Removing decimal test that doesn't work with Scala 2.11
f5cad4e [Patrick Wendell] Add Scala 2.11 docs
210d7e1 [Patrick Wendell] Revert "Testing new Hive version with shaded jline"
48518ce [Patrick Wendell] Remove association of Hive and Thriftserver profiles.
e9d0a06 [Patrick Wendell] Revert "Enable thritfserver for Scala 2.10 only"
67ec364 [Patrick Wendell] Guard building of thriftserver around Scala 2.10 check
8502c23 [Patrick Wendell] Enable thritfserver for Scala 2.10 only
e22b104 [Patrick Wendell] Small fix in pom file
ec402ab [Patrick Wendell] Various fixes
0be5a9d [Patrick Wendell] Testing new Hive version with shaded jline
4eaec65 [Prashant Sharma] Changed scripts to ignore target.
5167bea [Prashant Sharma] small correction
a4fcac6 [Prashant Sharma] Run against scala 2.11 on jenkins.
80285f4 [Prashant Sharma] MAven equivalent of setting spark.executor.extraClasspath during tests.
034b369 [Prashant Sharma] Setting test jars on executor classpath during tests from sbt.
d4874cb [Prashant Sharma] Fixed Python Runner suite. null check should be first case in scala 2.11.
6f50f13 [Prashant Sharma] Fixed build after rebasing with master. We should use ${scala.binary.version} instead of just 2.10
e56ca9d [Prashant Sharma] Print an error if build for 2.10 and 2.11 is spotted.
937c0b8 [Prashant Sharma] SCALA_VERSION -> SPARK_SCALA_VERSION
cb059b0 [Prashant Sharma] Code review
0476e5e [Prashant Sharma] Scala 2.11 support with repl and all build changes.
Author: Timothy Chen <tnachen@gmail.com>
Closes#1487 from tnachen/resource_offer_refactor and squashes the following commits:
4ea5dec [Timothy Chen] Rebase from master and address comments
9ccab09 [Timothy Chen] Address review comments
e6494dc [Timothy Chen] Refactor class loading
8207428 [Timothy Chen] Refactor mesos scheduler resourceOffers and add unit test
Netty's DefaultFileRegion requires a FileDescriptor in its constructor, which means we need to have a opened file handle. In super large workloads, this could lead to too many open files due to the way these file descriptors are cleaned. This pull request creates a new LazyFileRegion that initializes the FileDescriptor when we are sending data for the first time.
Author: Reynold Xin <rxin@databricks.com>
Author: Reynold Xin <rxin@apache.org>
Closes#3172 from rxin/lazyFD and squashes the following commits:
0bdcdc6 [Reynold Xin] Added reference to Netty's DefaultFileRegion
d4564ae [Reynold Xin] Added SparkConf to the ctor argument of IndexShuffleBlockManager.
6ed369e [Reynold Xin] Code review feedback.
04cddc8 [Reynold Xin] [SPARK-4307] Initialize FileDescriptor lazily in FileRegion.
For me the core tests failed because there are two locale dependent parts in the code.
Look at the Jira ticket for details.
Why is it necessary to check the exception message in isBindCollision in
https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/util/Utils.scala#L1686
?
Author: Niklas Wilcke <1wilcke@informatik.uni-hamburg.de>
Closes#3036 from numbnut/core-test-fix and squashes the following commits:
1fb0d04 [Niklas Wilcke] Fixing locale dependend code and tests
Author: Sandy Ryza <sandy@cloudera.com>
This patch had conflicts when merged, resolved by
Committer: Kay Ousterhout <kayousterhout@gmail.com>
Closes#2968 from sryza/sandy-spark-3179 and squashes the following commits:
dce4784 [Sandy Ryza] More review feedback
8d350d1 [Sandy Ryza] Fix test against Hadoop 2.5+
e7c74d0 [Sandy Ryza] More review feedback
6cff9c4 [Sandy Ryza] Review feedback
fb2dde0 [Sandy Ryza] SPARK-3179
andrewor14 Another try at SPARK-1209, to address https://github.com/apache/spark/pull/2814#issuecomment-61197619
I successfully tested with `mvn -Dhadoop.version=1.0.4 -DskipTests clean package; mvn -Dhadoop.version=1.0.4 test` I assume that is what failed Jenkins last time. I also tried `-Dhadoop.version1.2.1` and `-Phadoop-2.4 -Pyarn -Phive` for more coverage.
So this is why the class was put in `org.apache.hadoop` to begin with, I assume. One option is to leave this as-is for now and move it only when Hadoop 1.0.x support goes away.
This is the other option, which adds a call to force the constructor to be public at run-time. It's probably less surprising than putting Spark code in `org.apache.hadoop`, but, does involve reflection. A `SecurityManager` might forbid this, but it would forbid a lot of stuff Spark does. This would also only affect Hadoop 1.0.x it seems.
Author: Sean Owen <sowen@cloudera.com>
Closes#3048 from srowen/SPARK-1209 and squashes the following commits:
0d48f4b [Sean Owen] For Hadoop 1.0.x, make certain constructors public, which were public in later versions
466e179 [Sean Owen] Disable MIMA warnings resulting from moving the class -- this was also part of the PairRDDFunctions type hierarchy though?
eb61820 [Sean Owen] Move SparkHadoopMapRedUtil / SparkHadoopMapReduceUtil from org.apache.hadoop to org.apache.spark
Use "k" in javadoc of top and takeOrdered to avoid confusion with type K in pair RDDs. I think this resolves the discussion in SPARK-1344.
Author: Sean Owen <sowen@cloudera.com>
Closes#3168 from srowen/SPARK-1344 and squashes the following commits:
6963fcc [Sean Owen] Use "k" in javadoc of top and takeOrdered to avoid confusion with type K in pair RDDs
This PR resorts to `SparkContext.version` rather than META-INF/MANIFEST.MF in the assembly jar to inspect Spark version. Currently, when built with Maven, the MANIFEST.MF file in the assembly jar is incorrectly replaced by Guava 15.0 MANIFEST.MF, probably because of the assembly/shading tricks.
Another related PR is #3103, which tries to fix the MANIFEST issue.
Author: Cheng Lian <lian@databricks.com>
Closes#3105 from liancheng/spark-4225 and squashes the following commits:
d9585e1 [Cheng Lian] Resorts to SparkContext.version to inspect Spark version
This PR elimiantes the network package's usage of the Java serializer and replaces it with Encodable, which is a lightweight binary protocol. Each message is preceded by a type id, which will allow us to change messages (by only adding new ones), or to change the format entirely by switching to a special id (such as -1).
This protocol has the advantage over Java that we can guarantee that messages will remain compatible across compiled versions and JVMs, though it does not provide a clean way to do schema migration. In the future, it may be good to use a more heavy-weight serialization format like protobuf, thrift, or avro, but these all add several dependencies which are unnecessary at the present time.
Additionally this unifies the RPC messages of NettyBlockTransferService and ExternalShuffleClient.
Author: Aaron Davidson <aaron@databricks.com>
Closes#3146 from aarondav/free and squashes the following commits:
ed1102a [Aaron Davidson] Remove some unused imports
b8e2a49 [Aaron Davidson] Add appId to test
538f2a3 [Aaron Davidson] [SPARK-4187] [Core] Switch to binary protocol for external shuffle service messages
This PR fixed `Utils.exceptionString` to output the full exception information. However, the stack trace may become very huge, so I also updated the Web UI to collapse the error information by default (display the first line and clicking `+detail` will display the full info).
Here are the screenshots:
Stages:
Details for one stage:
The full information in the gray text field is:
```Java
org.apache.spark.shuffle.FetchFailedException: Connection reset by peer
at org.apache.spark.shuffle.hash.BlockStoreShuffleFetcher$.org$apache$spark$shuffle$hash$BlockStoreShuffleFetcher$$unpackBlock$1(BlockStoreShuffleFetcher.scala:67)
at org.apache.spark.shuffle.hash.BlockStoreShuffleFetcher$$anonfun$3.apply(BlockStoreShuffleFetcher.scala:83)
at org.apache.spark.shuffle.hash.BlockStoreShuffleFetcher$$anonfun$3.apply(BlockStoreShuffleFetcher.scala:83)
at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:371)
at org.apache.spark.util.CompletionIterator.hasNext(CompletionIterator.scala:30)
at org.apache.spark.InterruptibleIterator.hasNext(InterruptibleIterator.scala:39)
at scala.collection.Iterator$$anon$11.hasNext(Iterator.scala:327)
at scala.collection.Iterator$$anon$11.hasNext(Iterator.scala:327)
at org.apache.spark.util.collection.ExternalAppendOnlyMap.insertAll(ExternalAppendOnlyMap.scala:129)
at org.apache.spark.rdd.CoGroupedRDD$$anonfun$compute$5.apply(CoGroupedRDD.scala:160)
at org.apache.spark.rdd.CoGroupedRDD$$anonfun$compute$5.apply(CoGroupedRDD.scala:159)
at scala.collection.TraversableLike$WithFilter$$anonfun$foreach$1.apply(TraversableLike.scala:772)
at scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:47)
at scala.collection.TraversableLike$WithFilter.foreach(TraversableLike.scala:771)
at org.apache.spark.rdd.CoGroupedRDD.compute(CoGroupedRDD.scala:159)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:263)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:230)
at org.apache.spark.rdd.MappedValuesRDD.compute(MappedValuesRDD.scala:31)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:263)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:230)
at org.apache.spark.rdd.FlatMappedValuesRDD.compute(FlatMappedValuesRDD.scala:31)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:263)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:230)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:61)
at org.apache.spark.scheduler.Task.run(Task.scala:56)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:189)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:662)
Caused by: java.io.IOException: Connection reset by peer
at sun.nio.ch.FileDispatcher.read0(Native Method)
at sun.nio.ch.SocketDispatcher.read(SocketDispatcher.java:21)
at sun.nio.ch.IOUtil.readIntoNativeBuffer(IOUtil.java:198)
at sun.nio.ch.IOUtil.read(IOUtil.java:166)
at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:245)
at io.netty.buffer.PooledUnsafeDirectByteBuf.setBytes(PooledUnsafeDirectByteBuf.java:311)
at io.netty.buffer.AbstractByteBuf.writeBytes(AbstractByteBuf.java:881)
at io.netty.channel.socket.nio.NioSocketChannel.doReadBytes(NioSocketChannel.java:225)
at io.netty.channel.nio.AbstractNioByteChannel$NioByteUnsafe.read(AbstractNioByteChannel.java:119)
at io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:511)
at io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:468)
at io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:382)
at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:354)
at io.netty.util.concurrent.SingleThreadEventExecutor$2.run(SingleThreadEventExecutor.java:116)
... 1 more
```
/cc aarondav
Author: zsxwing <zsxwing@gmail.com>
Closes#3073 from zsxwing/SPARK-4204 and squashes the following commits:
176d1e3 [zsxwing] Add comments to explain the stack trace difference
ca509d3 [zsxwing] Add fullStackTrace to the constructor of ExceptionFailure
a07057b [zsxwing] Core style fix
dfb0032 [zsxwing] Backward compatibility for old history server
1e50f71 [zsxwing] Update as per review and increase the max height of the stack trace details
94f2566 [zsxwing] Change Utils.exceptionString to contain the inner exceptions and make the error information in Web UI more friendly
This relies on a hook from whoever is hosting the shuffle service to invoke removeApplication() when the application is completed. Once invoked, we will clean up all the executors' shuffle directories we know about.
Author: Aaron Davidson <aaron@databricks.com>
Closes#3126 from aarondav/cleanup and squashes the following commits:
33a64a9 [Aaron Davidson] Missing brace
e6e428f [Aaron Davidson] Address comments
16a0d27 [Aaron Davidson] Cleanup
e4df3e7 [Aaron Davidson] [SPARK-4236] Cleanup removed applications' files in shuffle service
This adds a RetryingBlockFetcher to the NettyBlockTransferService which is wrapped around our typical OneForOneBlockFetcher, adding retry logic in the event of an IOException.
This sort of retry allows us to avoid marking an entire executor as failed due to garbage collection or high network load.
TODO:
- [x] unit tests
- [x] put in ExternalShuffleClient too
Author: Aaron Davidson <aaron@databricks.com>
Closes#3101 from aarondav/retry and squashes the following commits:
72a2a32 [Aaron Davidson] Add that we should remove the condition around the retry thingy
c7fd107 [Aaron Davidson] Fix unit tests
e80e4c2 [Aaron Davidson] Address initial comments
6f594cd [Aaron Davidson] Fix unit test
05ff43c [Aaron Davidson] Add to external shuffle client and add unit test
66e5a24 [Aaron Davidson] [SPARK-4238] [Core] Perform network-level retry of shuffle file fetches
Author: Aaron Davidson <aaron@databricks.com>
Closes#3142 from aarondav/worker and squashes the following commits:
3780bd7 [Aaron Davidson] Address comments
2dcdfc1 [Aaron Davidson] Add private[worker]
47f49d3 [Aaron Davidson] NettyBlockTransferService shouldn't care about app ids (it's only b/t executors)
258417c [Aaron Davidson] [SPARK-4277] Support external shuffle service on executor
Author: Aaron Davidson <aaron@databricks.com>
Closes#3128 from aarondav/compiter and squashes the following commits:
698e4be [Aaron Davidson] [SPARK-4264] Completion iterator should only invoke callback once
add binaryFiles() and binaryRecords() in Python
```
binaryFiles(self, path, minPartitions=None):
:: Developer API ::
Read a directory of binary files from HDFS, a local file system
(available on all nodes), or any Hadoop-supported file system URI
as a byte array. Each file is read as a single record and returned
in a key-value pair, where the key is the path of each file, the
value is the content of each file.
Note: Small files are preferred, large file is also allowable, but
may cause bad performance.
binaryRecords(self, path, recordLength):
Load data from a flat binary file, assuming each record is a set of numbers
with the specified numerical format (see ByteBuffer), and the number of
bytes per record is constant.
:param path: Directory to the input data files
:param recordLength: The length at which to split the records
```
Author: Davies Liu <davies@databricks.com>
Closes#3078 from davies/binary and squashes the following commits:
cd0bdbd [Davies Liu] Merge branch 'master' of github.com:apache/spark into binary
3aa349b [Davies Liu] add experimental notes
24e84b6 [Davies Liu] Merge branch 'master' of github.com:apache/spark into binary
5ceaa8a [Davies Liu] Merge branch 'master' of github.com:apache/spark into binary
1900085 [Davies Liu] bugfix
bb22442 [Davies Liu] add binaryFiles and binaryRecords in Python
This commit stripes table rows after hiding some rows, to
ensure that rows are correct striped to alternate white
and grey even when rows are hidden by default.
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#3117 from kayousterhout/striping and squashes the following commits:
be6e10a [Kay Ousterhout] [SPARK-4255] Fix incorrect table striping
Mesos offers may not contain all resources, and Spark needs to check to
ensure they are present and sufficient. Spark may throw an erroneous
exception when resources aren't present.
Author: Brenden Matthews <brenden@diddyinc.com>
Closes#3024 from brndnmtthws/fix-mesos-resource-misuse and squashes the following commits:
e5f9580 [Brenden Matthews] [SPARK-4158] Fix for missing resources.
This creates a new module `network/yarn` that depends on `network/shuffle` recently created in #3001. This PR introduces a custom Yarn auxiliary service that runs the external shuffle service. As of the changes here this shuffle service is required for using dynamic allocation with Spark.
This is still WIP mainly because it doesn't handle security yet. I have tested this on a stable Yarn cluster.
Author: Andrew Or <andrew@databricks.com>
Closes#3082 from andrewor14/yarn-shuffle-service and squashes the following commits:
ef3ddae [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-shuffle-service
0ee67a2 [Andrew Or] Minor wording suggestions
1c66046 [Andrew Or] Remove unused provided dependencies
0eb6233 [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-shuffle-service
6489db5 [Andrew Or] Try catch at the right places
7b71d8f [Andrew Or] Add detailed java docs + reword a few comments
d1124e4 [Andrew Or] Add security to shuffle service (INCOMPLETE)
5f8a96f [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-shuffle-service
9b6e058 [Andrew Or] Address various feedback
f48b20c [Andrew Or] Fix tests again
f39daa6 [Andrew Or] Do not make network-yarn an assembly module
761f58a [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-shuffle-service
15a5b37 [Andrew Or] Fix build for Hadoop 1.x
baff916 [Andrew Or] Fix tests
5bf9b7e [Andrew Or] Address a few minor comments
5b419b8 [Andrew Or] Add missing license header
804e7ff [Andrew Or] Include the Yarn shuffle service jar in the distribution
cd076a4 [Andrew Or] Require external shuffle service for dynamic allocation
ea764e0 [Andrew Or] Connect to Yarn shuffle service only if it's enabled
1bf5109 [Andrew Or] Use the shuffle service port specified through hadoop config
b4b1f0c [Andrew Or] 4 tabs -> 2 tabs
43dcb96 [Andrew Or] First cut integration of shuffle service with Yarn aux service
b54a0c4 [Andrew Or] Initial skeleton for Yarn shuffle service
replaces the existing read() call with readFully().
Author: industrial-sloth <industrial-sloth@users.noreply.github.com>
Closes#3093 from industrial-sloth/branch-1.2-fixedLenRecRdr and squashes the following commits:
a245c8a [industrial-sloth] use readFully in FixedLengthBinaryRecordReader
(cherry picked from commit 6844e7a821)
Signed-off-by: Matei Zaharia <matei@databricks.com>
This commit fixes the scheduler delay in the UI (which previously
included things that are not scheduler delay, like time to
deserialize the task and serialize the result), and also
adds information about time to deserialize tasks to the optional
additional metrics. Time to deserialize the task can be large relative
to task time for short jobs, and understanding when it is high can help
developers realize that they should try to reduce closure size (e.g, by including
less data in the task description).
cc shivaram etrain
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#2832 from kayousterhout/SPARK-3983 and squashes the following commits:
0c1398e [Kay Ousterhout] Fixed ordering
531575d [Kay Ousterhout] Removed executor launch time
1f13afe [Kay Ousterhout] Minor spacing fixes
335be4b [Kay Ousterhout] Made metrics hideable
5bc3cba [Kay Ousterhout] [SPARK-3984] [SPARK-3983] Improve UI task metrics.
Does three things: (1) Adds SASL to ExternalShuffleClient, (2) puts SecurityManager in BlockManager's constructor, and (3) adds unit test.
Author: Aaron Davidson <aaron@databricks.com>
Closes#3108 from aarondav/sasl-client and squashes the following commits:
48b622d [Aaron Davidson] Screw it, let's just get LimitedInputStream
3543b70 [Aaron Davidson] Back out of pom change due to unknown test issue?
b58518a [Aaron Davidson] ByteStreams.limit() not available :(
cbe451a [Aaron Davidson] Address comments
2bf2908 [Aaron Davidson] [SPARK-4242] [Core] Add SASL to external shuffle service
This PR simplify serializer, always use batched serializer (AutoBatchedSerializer as default), even batch size is 1.
Author: Davies Liu <davies@databricks.com>
This patch had conflicts when merged, resolved by
Committer: Josh Rosen <joshrosen@databricks.com>
Closes#2920 from davies/fix_autobatch and squashes the following commits:
e544ef9 [Davies Liu] revert unrelated change
6880b14 [Davies Liu] Merge branch 'master' of github.com:apache/spark into fix_autobatch
1d557fc [Davies Liu] fix tests
8180907 [Davies Liu] Merge branch 'master' of github.com:apache/spark into fix_autobatch
76abdce [Davies Liu] clean up
53fa60b [Davies Liu] Merge branch 'master' of github.com:apache/spark into fix_autobatch
d7ac751 [Davies Liu] Merge branch 'master' of github.com:apache/spark into fix_autobatch
2cc2497 [Davies Liu] Merge branch 'master' of github.com:apache/spark into fix_autobatch
b4292ce [Davies Liu] fix bug in master
d79744c [Davies Liu] recover hive tests
be37ece [Davies Liu] refactor
eb3938d [Davies Liu] refactor serializer in scala
8d77ef2 [Davies Liu] simplify serializer, use AutoBatchedSerializer by default.
This patch allows executor thread dumps to be collected on-demand and viewed in the Spark web UI.
The thread dumps are collected using Thread.getAllStackTraces(). To allow remote thread dumps to be triggered from the web UI, I added a new `ExecutorActor` that runs inside of the Executor actor system and responds to RPCs from the driver. The driver's mechanism for obtaining a reference to this actor is a little bit hacky: it uses the block manager master actor to determine the host/port of the executor actor systems in order to construct ActorRefs to ExecutorActor. Unfortunately, I couldn't find a much cleaner way to do this without a big refactoring of the executor -> driver communication.
Screenshots:
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2944 from JoshRosen/jstack-in-web-ui and squashes the following commits:
3c21a5d [Josh Rosen] Address review comments:
880f7f7 [Josh Rosen] Merge remote-tracking branch 'origin/master' into jstack-in-web-ui
f719266 [Josh Rosen] Merge remote-tracking branch 'origin/master' into jstack-in-web-ui
19707b0 [Josh Rosen] Add one comment.
127a130 [Josh Rosen] Update to use SparkContext.DRIVER_IDENTIFIER
b8e69aa [Josh Rosen] Merge remote-tracking branch 'origin/master' into jstack-in-web-ui
3dfc2d4 [Josh Rosen] Add missing file.
bc1e675 [Josh Rosen] Undo some leftover changes from the earlier approach.
f4ac1c1 [Josh Rosen] Switch to on-demand collection of thread dumps
dfec08b [Josh Rosen] Add option to disable thread dumps in UI.
4c87d7f [Josh Rosen] Use separate RPC for sending thread dumps.
2b8bdf3 [Josh Rosen] Enable thread dumps from the driver when running in non-local mode.
cc3e6b3 [Josh Rosen] Fix test code in DAGSchedulerSuite.
87b8b65 [Josh Rosen] Add new listener event for thread dumps.
8c10216 [Josh Rosen] Add missing file.
0f198ac [Josh Rosen] [SPARK-611] Display executor thread dumps in web UI
The number of completed stages and failed stages showed on webUI will always be less than 1000. This is really misleading when there are already thousands of stages completed or failed. The number should be correct even when only partial stages listed on the webUI (stage info will be removed if the number is too large).
Author: Zhang, Liye <liye.zhang@intel.com>
Closes#3035 from liyezhang556520/webStageNum and squashes the following commits:
d9e29fb [Zhang, Liye] add detailed comments for variables
4ea8fd1 [Zhang, Liye] change variable name accroding to comments
f4c404d [Zhang, Liye] [SPARK-4168][WebUI] web statges number should show correctly when stages are more than 1000
A leak of event loops may be causing test failures.
Author: Aaron Davidson <aaron@databricks.com>
Closes#3053 from aarondav/leak and squashes the following commits:
e676d18 [Aaron Davidson] Typo!
8f96475 [Aaron Davidson] Keep original ssc semantics
7e49f10 [Aaron Davidson] A leak of event loops may be causing test failures.
This PR overrides the `GetInfo` Hive Thrift API to provide correct version information. Another property `spark.sql.hive.version` is added to reveal the underlying Hive version. These are generally useful for Spark SQL ODBC driver providers. The Spark version information is extracted from the jar manifest. Also took the chance to remove the `SET -v` hack, which was a workaround for Simba ODBC driver connectivity.
TODO
- [x] Find a general way to figure out Hive (or even any dependency) version.
This [blog post](http://blog.soebes.de/blog/2014/01/02/version-information-into-your-appas-with-maven/) suggests several methods to inspect application version. In the case of Spark, this can be tricky because the chosen method:
1. must applies to both Maven build and SBT build
For Maven builds, we can retrieve the version information from the META-INF/maven directory within the assembly jar. But this doesn't work for SBT builds.
2. must not rely on the original jars of dependencies to extract specific dependency version, because Spark uses assembly jar.
This implies we can't read Hive version from Hive jar files since standard Spark distribution doesn't include them.
3. should play well with `SPARK_PREPEND_CLASSES` to ease local testing during development.
`SPARK_PREPEND_CLASSES` prevents classes to be loaded from the assembly jar, thus we can't locate the jar file and read its manifest.
Given these, maybe the only reliable method is to generate a source file containing version information at build time. pwendell Do you have any suggestions from the perspective of the build process?
**Update** Hive version is now retrieved from the newly introduced `HiveShim` object.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Author: Cheng Lian <lian@databricks.com>
Closes#2843 from liancheng/get-info and squashes the following commits:
a873d0f [Cheng Lian] Updates test case
53f43cd [Cheng Lian] Retrieves underlying Hive verson via HiveShim
1d282b8 [Cheng Lian] Removes the Simba ODBC "SET -v" hack
f857fce [Cheng Lian] Overrides Hive GetInfo Thrift API and adds Hive version property
Now when ExecutorLostFailure happens, it only displays `ExecutorLostFailure (executor lost)`. Adding the executor id will help locate the faulted executor.
Author: zsxwing <zsxwing@gmail.com>
Closes#3033 from zsxwing/SPARK-4166 and squashes the following commits:
ff4664c [zsxwing] Backward-compatible support
c5c4cf2 [zsxwing] Display the executor ID in the Web UI when ExecutorLostFailure happens
Right now, operations like collect() and take() can crash the driver with an OOM if they bring back too many data.
This PR will introduce spark.driver.maxResultSize, after setting it, the driver will abort a job if its result is bigger than it.
By default, it's 1g (for backward compatibility for most the cases).
In local mode, the driver and executor share the same JVM, the default setting can not protect JVM from OOM.
cc mateiz
Author: Davies Liu <davies@databricks.com>
Closes#3003 from davies/collect and squashes the following commits:
248ed5e [Davies Liu] fix compile
272522e [Davies Liu] address comments
2c35773 [Davies Liu] add sizes in message of abort()
5d62303 [Davies Liu] address comments
bc3c077 [Davies Liu] Merge branch 'master' of github.com:apache/spark into collect
11f97c5 [Davies Liu] address comments
47b144f [Davies Liu] check the size of result before send and fetch
3d81af2 [Davies Liu] address comments
ca8267d [Davies Liu] limit the size of data by collect
This patch introduces the tooling necessary to construct an external shuffle service which is independent of Spark executors, and then use this service inside Spark. An example (just for the sake of this PR) of the service creation can be found in Worker, and the service itself is used by plugging in the StandaloneShuffleClient as Spark's ShuffleClient (setup in BlockManager).
This PR continues the work from #2753, which extracted out the transport layer of Spark's block transfer into an independent package within Spark. A new package was created which contains the Spark business logic necessary to retrieve the actual shuffle data, which is completely independent of the transport layer introduced in the previous patch. Similar to the transport layer, this package must not depend on Spark as we anticipate plugging this service as a lightweight process within, say, the YARN NodeManager, and do not wish to include Spark's dependencies (including Scala itself).
There are several outstanding tasks which must be complete before this PR can be merged:
- [x] Complete unit testing of network/shuffle package.
- [x] Performance and correctness testing on a real cluster.
- [x] Remove example service instantiation from Worker.scala.
There are even more shortcomings of this PR which should be addressed in followup patches:
- Don't use Java serializer for RPC layer! It is not cross-version compatible.
- Handle shuffle file cleanup for dead executors once the application terminates or the ContextCleaner triggers.
- Documentation of the feature in the Spark docs.
- Improve behavior if the shuffle service itself goes down (right now we don't blacklist it, and new executors cannot spawn on that machine).
- SSL and SASL integration
- Nice to have: Handle shuffle file consolidation (this would requires changes to Spark's implementation).
Author: Aaron Davidson <aaron@databricks.com>
Closes#3001 from aarondav/shuffle-service and squashes the following commits:
4d1f8c1 [Aaron Davidson] Remove changes to Worker
705748f [Aaron Davidson] Rename Standalone* to External*
fd3928b [Aaron Davidson] Do not unregister executor outputs unduly
9883918 [Aaron Davidson] Make suggested build changes
3d62679 [Aaron Davidson] Add Spark integration test
7fe51d5 [Aaron Davidson] Fix SBT integration
56caa50 [Aaron Davidson] Address comments
c8d1ac3 [Aaron Davidson] Add unit tests
2f70c0c [Aaron Davidson] Fix unit tests
5483e96 [Aaron Davidson] Fix unit tests
46a70bf [Aaron Davidson] Whoops, bracket
5ea4df6 [Aaron Davidson] [SPARK-3796] Create external service which can serve shuffle files
Note that we're turning this on for at least the first part of the QA period as a trial. We want to enable this (and deprecate the NioBlockTransferService) as soon as possible in the hopes that NettyBlockTransferService will be more stable and easier to maintain. We will turn it off if we run into major issues.
Author: Aaron Davidson <aaron@databricks.com>
Closes#3049 from aarondav/enable-netty and squashes the following commits:
bb981cc [Aaron Davidson] [SPARK-4183] Enable NettyBlockTransferService by default
The additions add the abstract BinaryFileInputFormat and BinaryRecordReader classes for reading in data as a byte stream and converting it to another format using the ```def parseByteArray(inArray: Array[Byte]): T``` function.
As a trivial example ```ByteInputFormat``` and ```ByteRecordReader``` are included which just return the Array[Byte] from a given file.
Finally a RDD for ```BinaryFileInputFormat``` (to allow for easier partitioning changes as was done for WholeFileInput) was added and the appropriate byteFiles to the ```SparkContext``` so the functions can be easily used by others.
A common use case might be to read in a folder
```
sc.byteFiles("s3://mydrive/tif/*.tif").map(rawData => ReadTiffFromByteArray(rawData))
```
Author: Kevin Mader <kevinmader@gmail.com>
Author: Kevin Mader <kmader@users.noreply.github.com>
Closes#1658 from kmader/master and squashes the following commits:
3c49a30 [Kevin Mader] fixing wholetextfileinput to it has the same setMinPartitions function as in BinaryData files
359a096 [Kevin Mader] making the final corrections suggested by @mateiz and renaming a few functions to make their usage clearer
6379be4 [Kevin Mader] reorganizing code
7b9d181 [Kevin Mader] removing developer API, cleaning up imports
8ac288b [Kevin Mader] fixed a single slightly over 100 character line
92bda0d [Kevin Mader] added new tests, renamed files, fixed several of the javaapi functions, formatted code more nicely
a32fef7 [Kevin Mader] removed unneeded classes added DeveloperApi note to portabledatastreams since the implementation might change
49174d9 [Kevin Mader] removed unneeded classes added DeveloperApi note to portabledatastreams since the implementation might change
c27a8f1 [Kevin Mader] jenkins crashed before running anything last time, so making minor change
b348ce1 [Kevin Mader] fixed order in check (prefix only appears on jenkins not when I run unit tests locally)
0588737 [Kevin Mader] filename check in "binary file input as byte array" test now ignores prefixes and suffixes which might get added by Hadoop
4163e38 [Kevin Mader] fixing line length and output from FSDataInputStream to DataInputStream to minimize sensitivity to Hadoop API changes
19812a8 [Kevin Mader] Fixed the serialization issue with PortableDataStream since neither CombineFileSplit nor TaskAttemptContext implement the Serializable interface, by using ByteArrays for storing both and then recreating the objects from these bytearrays as needed.
238c83c [Kevin Mader] fixed several scala-style issues, changed structure of binaryFiles, removed excessive classes added new tests. The caching tests still have a serialization issue, but that should be easily fixed as well.
932a206 [Kevin Mader] Update RawFileInput.scala
a01c9cf [Kevin Mader] Update RawFileInput.scala
441f79a [Kevin Mader] fixed a few small comments and dependency
12e7be1 [Kevin Mader] removing imglib from maven (definitely not ready yet)
5deb79e [Kevin Mader] added new portabledatastream to code so that it can be serialized correctly
f032bc0 [Kevin Mader] fixed bug in path name, renamed tests
bc5c0b9 [Kevin Mader] made minor stylistic adjustments from mateiz
df8e528 [Kevin Mader] fixed line lengths and changed java test
9a313d5 [Kevin Mader] making classes that needn't be public private, adding automatic file closure, adding new tests
edf5829 [Kevin Mader] fixing line lengths, adding new lines
f4841dc [Kevin Mader] un-optimizing imports, silly intellij
eacfaa6 [Kevin Mader] Added FixedLengthBinaryInputFormat and RecordReader from freeman-lab and added them to both the JavaSparkContext and the SparkContext as fixedLengthBinaryFile
1622935 [Kevin Mader] changing the line lengths to make jenkins happy
1cfa38a [Kevin Mader] added apache headers, added datainputstream directly as an output option for more complicated readers (HDF5 perhaps), and renamed several of the functions and files to be more consistent. Also added parallel functions to the java api
84035f1 [Kevin Mader] adding binary and byte file support spark
81c5f12 [Kevin Mader] Merge pull request #1 from apache/master
Author: Sandy Ryza <sandy@cloudera.com>
Closes#3043 from sryza/sandy-spark-4175 and squashes the following commits:
e327340 [Sandy Ryza] SPARK-4175. Exception on stage page
WebUI
Author: Mark Mims <mark.mims@canonical.com>
This patch had conflicts when merged, resolved by
Committer: Josh Rosen <joshrosen@databricks.com>
Closes#3031 from mmm/remove-accumulators-col and squashes the following commits:
6141cb3 [Mark Mims] reformat to satisfy scalastyle linelength. build failed from jenkins https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/22604/
390893b [Mark Mims] cleanup
c28c449 [Mark Mims] looking much better now... minimal explicit formatting. Now, see if any sort keys make sense
fb72156 [Mark Mims] mimic hasInput. The basics work here, but wanna clean this up with maybeAccumulators for column content
This commit adds a set of checkboxes to the stage detail
page that the user can use to show additional task metrics,
including the GC time, result serialization time, result fetch
time, and scheduler delay. All of these metrics are now
hidden by default. This allows advanced users to look at more
detailed metrics, without distracting the average user.
This change also cleans up the stage detail page so that metrics
are shown in the same order in the summary table as in the task table,
and updates the metrics in both tables such that they contain the same
set of metrics.
The ability to remember a user's preferences for which metrics
should be shown has been filed as SPARK-4024.
Here's what the stage detail page looks like by default:
and once a user clicks "Show additional metrics" (note that all the metrics get checked by default):
cc shivaram andrewor14
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#2867 from kayousterhout/SPARK-checkboxes and squashes the following commits:
6015913 [Kay Ousterhout] Added comment
08dee73 [Kay Ousterhout] Josh's usability comments
0940d61 [Kay Ousterhout] Style updates based on Andrew's review
ef05ccd [Kay Ousterhout] Added tooltips
d7cfaaf [Kay Ousterhout] Made list of add'l metrics collapsible.
70c1fb5 [Kay Ousterhout] [SPARK-4016] Allow user to show/hide UI metrics.
More efficient sampling, based on Gap Sampling optimization:
http://erikerlandson.github.io/blog/2014/09/11/faster-random-samples-with-gap-sampling/
Author: Erik Erlandson <eerlands@redhat.com>
Closes#2455 from erikerlandson/spark-3250-pr and squashes the following commits:
72496bc [Erik Erlandson] [SPARK-3250] Implement Gap Sampling optimization for random sampling
(This is just a look at what completely moving the classes would look like. I know Patrick flagged that as maybe not OK, although, it's private?)
Author: Sean Owen <sowen@cloudera.com>
Closes#2814 from srowen/SPARK-1209 and squashes the following commits:
ead1115 [Sean Owen] Disable MIMA warnings resulting from moving the class -- this was also part of the PairRDDFunctions type hierarchy though?
2d42c1d [Sean Owen] Move SparkHadoopMapRedUtil / SparkHadoopMapReduceUtil from org.apache.hadoop to org.apache.spark
This also includes minor re-organization of the code. Tested locally in both client and deploy modes.
Author: Andrew Or <andrew@databricks.com>
Author: Andrew Or <andrewor14@gmail.com>
Closes#2697 from andrewor14/memory-cluster-mode and squashes the following commits:
01d78bc [Andrew Or] Merge branch 'master' of github.com:apache/spark into memory-cluster-mode
ccd468b [Andrew Or] Add some comments per Patrick
c956577 [Andrew Or] Tweak wording
2b4afa0 [Andrew Or] Unused import
47a5a88 [Andrew Or] Correct Spark properties precedence order
bf64717 [Andrew Or] Merge branch 'master' of github.com:apache/spark into memory-cluster-mode
dd452d0 [Andrew Or] Respect spark.*.memory in cluster mode
Sort "Started", "Completed", "Duration" and "Last Updated" by time.
Author: zsxwing <zsxwing@gmail.com>
Closes#3014 from zsxwing/SPARK-4153 and squashes the following commits:
ec8b9ad [zsxwing] Sort "Started", "Completed", "Duration" and "Last Updated" by time
I had planned to fix this when I merged it but I forgot to. witgo
Author: Andrew Or <andrew@databricks.com>
Closes#3018 from andrewor14/command-utils-style and squashes the following commits:
c2959fb [Andrew Or] Style hot fix
We use "\<driver\>" everywhere. Let's not do that.
Author: Andrew Or <andrew@databricks.com>
Closes#3020 from andrewor14/consolidate-driver and squashes the following commits:
c1c2204 [Andrew Or] Just use "<driver>" for local executor ID
3d751e9 [Andrew Or] Consolidate usages of <driver>
Author: Andrew Or <andrewor14@gmail.com>
Author: Andrew Or <andrew@databricks.com>
Closes#3021 from andrewor14/typos and squashes the following commits:
daaf417 [Andrew Or] Merge branch 'master' of github.com:apache/spark into typos
4838ae4 [Andrew Or] Merge branch 'master' of github.com:apache/spark into typos
026d426 [Andrew Or] Merge branch 'master' of github.com:andrewor14/spark into typos
a81ae8f [Andrew Or] Some typos
The bulk of this PR is comprised of tests. All changes in functionality are made in `SparkSubmit.scala` (~20 lines).
**SPARK-3319.** There is currently a divergence in behavior when the user passes in additional jars through `--jars` and through setting `spark.jars` in the default properties file. The former will happily resolve the paths (e.g. convert `my.jar` to `file:/absolute/path/to/my.jar`), while the latter does not. We should resolve paths consistently in both cases. This also applies to the following pairs of command line arguments and Spark configs:
- `--jars` ~ `spark.jars`
- `--files` ~ `spark.files` / `spark.yarn.dist.files`
- `--archives` ~ `spark.yarn.dist.archives`
- `--py-files` ~ `spark.submit.pyFiles`
**SPARK-3338.** This PR also fixes the following bug: if the user sets `spark.submit.pyFiles` in his/her properties file, it does not actually get picked up even if `--py-files` is not set. This is simply because the config is overridden by an empty string.
Author: Andrew Or <andrewor14@gmail.com>
Author: Andrew Or <andrew@databricks.com>
Closes#2232 from andrewor14/resolve-config-paths and squashes the following commits:
fff2869 [Andrew Or] Add spark.yarn.jar
da3a1c1 [Andrew Or] Merge branch 'master' of github.com:apache/spark into resolve-config-paths
f0fae64 [Andrew Or] Merge branch 'master' of github.com:apache/spark into resolve-config-paths
05e03d6 [Andrew Or] Add tests for resolving both command line and config paths
460117e [Andrew Or] Resolve config paths properly
fe039d3 [Andrew Or] Beef up tests to test fixed-pointed-ness of Utils.resolveURI(s)
By default, Spark builds its package against Hadoop 1.0.4 version. In that version, it has some FsPermission bug (see [HADOOP-7629] (https://issues.apache.org/jira/browse/HADOOP-7629) by Todd Lipcon). This bug got fixed since 1.1 version. By using that FsPermission.createImmutable() API, end-user may see some RPC exception like below (if turn on eventlog over HDFS). Here proposes a quick fix to avoid certain exception for all hadoop versions.
```
Exception in thread "main" java.io.IOException: Call to sr484/10.1.2.84:54310 failed on local exception: java.io.EOFException
at org.apache.hadoop.ipc.Client.wrapException(Client.java:1150)
at org.apache.hadoop.ipc.Client.call(Client.java:1118)
at org.apache.hadoop.ipc.RPC$Invoker.invoke(RPC.java:229)
at $Proxy6.setPermission(Unknown Source)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
at java.lang.reflect.Method.invoke(Method.java:597)
at org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:85)
at org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:62)
at $Proxy6.setPermission(Unknown Source)
at org.apache.hadoop.hdfs.DFSClient.setPermission(DFSClient.java:1285)
at org.apache.hadoop.hdfs.DistributedFileSystem.setPermission(DistributedFileSystem.java:572)
at org.apache.spark.util.FileLogger.createLogDir(FileLogger.scala:138)
at org.apache.spark.util.FileLogger.start(FileLogger.scala:115)
at org.apache.spark.scheduler.EventLoggingListener.start(EventLoggingListener.scala:74)
at org.apache.spark.SparkContext.<init>(SparkContext.scala:324)
```
Author: Grace <jie.huang@intel.com>
Closes#2892 from GraceH/eventlog-rpc and squashes the following commits:
58ea038 [Grace] new FsPermission Instance w/o FsPermission.createImmutable
As part of the initiative of preventing data loss on streaming driver failure, this sub-task implements a BlockRDD that is backed by HDFS. This BlockRDD can either read data from the Spark's BlockManager, or read the data from file-segments in write ahead log in HDFS.
Most of this code has been written by @harishreedharan
Author: Tathagata Das <tathagata.das1565@gmail.com>
Author: Hari Shreedharan <hshreedharan@apache.org>
Closes#2931 from tdas/driver-ha-rdd and squashes the following commits:
209e49c [Tathagata Das] Better fix to style issue.
4a5866f [Tathagata Das] Addressed one more comment.
ed5fbf0 [Tathagata Das] Minor updates.
b0a18b1 [Tathagata Das] Fixed import order.
20aa7c6 [Tathagata Das] Fixed more line length issues.
29aa099 [Tathagata Das] Fixed line length issues.
9e47b5b [Tathagata Das] Renamed class, simplified+added unit tests.
6e1bfb8 [Tathagata Das] Tweaks testuite to create spark contxt lazily to prevent contxt leaks.
9c86a61 [Tathagata Das] Merge pull request #22 from harishreedharan/driver-ha-rdd
2878c38 [Hari Shreedharan] Shutdown spark context after tests. Formatting/minor fixes
c709f2f [Tathagata Das] Merge pull request #21 from harishreedharan/driver-ha-rdd
5cce16f [Hari Shreedharan] Make sure getBlockLocations uses offset and length to find the blocks on HDFS
eadde56 [Tathagata Das] Transferred HDFSBackedBlockRDD for the driver-ha-working branch
This method is not implemented by the only subclass
(HashShuffleReader), nor is it ever called. While the
use of Scala's fancy "???" was pretty exciting, the method's
existence can only lead to confusion and it therefore should
be deleted.
mateiz was there a reason for adding this that I'm
missing?
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#2966 from kayousterhout/SPARK-4102 and squashes the following commits:
532c564 [Kay Ousterhout] Added back commented-out method, as per Matei's request
904655e [Kay Ousterhout] [SPARK-4102] Remove unused ShuffleReader.stop() method.
This is part of a bigger effort to provide elastic scaling of executors within a Spark application ([SPARK-3174](https://issues.apache.org/jira/browse/SPARK-3174)). This PR does not provide any functionality by itself; it is a skeleton that is missing a mechanism to be added later in [SPARK-3822](https://issues.apache.org/jira/browse/SPARK-3822).
Comments and feedback are most welcome. For those of you reviewing this in detail, I highly recommend doing it through your favorite IDE instead of through the diff here.
Author: Andrew Or <andrewor14@gmail.com>
Author: Andrew Or <andrew@databricks.com>
Closes#2746 from andrewor14/scaling-heuristics and squashes the following commits:
8a4fdaa [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
e045df8 [Andrew Or] Add warning message (minor)
dfa31ec [Andrew Or] Fix tests
c0becc4 [Andrew Or] Merging with SPARK-3822
4784f93 [Andrew Or] Reword an awkward log message
181f27f [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
c79e907 [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
4672b90 [Andrew Or] It's nano time.
a6a30f2 [Andrew Or] Do not allow min/max executors of 0
c60ec33 [Andrew Or] Rewrite test logic with clocks
b00b680 [Andrew Or] Fix style
c3caa65 [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
7f9da14 [Andrew Or] Factor out logic to verify bounds on # executors (minor)
f279019 [Andrew Or] Add time mocking tests for polling loop
685e347 [Andrew Or] Factor out clock in polling loop to facilitate testing
3cea7f7 [Andrew Or] Use PrivateMethodTester to keep original class private
3156d81 [Andrew Or] Update comments and exception messages
92f36f9 [Andrew Or] Address minor review comments
abdea61 [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
2aefd09 [Andrew Or] Correct listener behavior
9fe6e44 [Andrew Or] Rename variables and configs + update comments and log messages
149cc32 [Andrew Or] Fix style
254c958 [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
5ff829b [Andrew Or] Add tests for ExecutorAllocationManager
19c6c4b [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
5896515 [Andrew Or] Move ExecutorAllocationManager out of scheduler package
9ca8945 [Andrew Or] Rewrite callbacks through the listener interface
5e336b9 [Andrew Or] Remove code from backend to avoid conflict with SPARK-3822
092d1fd [Andrew Or] Remove timeout logic for pending requests
1309fab [Andrew Or] Request executors by specifying the number pending
8bc0e9d [Andrew Or] Add logic to expire pending requests after timeouts
b750ee1 [Andrew Or] Express timers in terms of expiration times + remove retry logic
7f8dd47 [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
9d516cc [Andrew Or] Bug fix: Actually trigger the add timer / add retry timer
44f1832 [Andrew Or] Rename configs to include time units
eaae7ef [Andrew Or] Address various review comments
6f8be6c [Andrew Or] Beef up comments on what each of the timers mean
baaa403 [Andrew Or] Simplify variable names (minor)
42beec8 [Andrew Or] Reset whether the add threshold is crossed on cancellation
9bcc0bc [Andrew Or] ExecutorScalingManager -> ExecutorAllocationManager
2784398 [Andrew Or] Merge branch 'master' of github.com:apache/spark into scaling-heuristics
5a97d9e [Andrew Or] Log retry attempts in INFO + clean up logging
2f55c9f [Andrew Or] Do not keep requesting executors even after max attempts
0acd1cb [Andrew Or] Rewrite timer logic with polling
b3c7d44 [Andrew Or] Start the retry timer for adding executors at the right time
9b5f2ea [Andrew Or] Wording changes in comments and log messages
c2203a5 [Andrew Or] Simplify code to access the scheduler backend
e519d08 [Andrew Or] Simplify initialization code
2cc87a7 [Andrew Or] Add retry logic for removing executors
d0b34a6 [Andrew Or] Add retry logic for adding executors
9cc4649 [Andrew Or] Simplifying synchronization logic
67c03c7 [Andrew Or] Correct semantics of adding executors + update comments
6c48ab0 [Andrew Or] Update synchronization comment
8901900 [Andrew Or] Simplify remove policy + change the semantics of add policy
1cc8444 [Andrew Or] Minor wording change
ae5b64a [Andrew Or] Add synchronization
20ec6b9 [Andrew Or] First cut implementation of removing executors dynamically
4077ae2 [Andrew Or] Minor code re-organization
6f1fa66 [Andrew Or] First cut implementation of adding executors dynamically
b2e6dcc [Andrew Or] Add skeleton interface for requesting / killing executors
There is a chance that `thread` is null when calling `thread.interrupt()`.
```Scala
override def cancel(): Unit = this.synchronized {
_cancelled = true
if (thread != null) {
thread.interrupt()
}
}
```
Should put `thread = null` into a `synchronized` block to fix the race condition.
Author: zsxwing <zsxwing@gmail.com>
Closes#2957 from zsxwing/SPARK-4097 and squashes the following commits:
edf0aee [zsxwing] Add comments to explain the lock
c5cfeca [zsxwing] Fix the race condition of 'thread'
This is part of a broader effort to enable dynamic scaling of executors ([SPARK-3174](https://issues.apache.org/jira/browse/SPARK-3174)). This is intended to work alongside SPARK-3795 (#2746), SPARK-3796 and SPARK-3797, but is functionally independently of these other issues.
The logic is built on top of PraveenSeluka's changes at #2798. This is different from the changes there in a few major ways: (1) the mechanism is implemented within the existing scheduler backend framework rather than in new `Actor` classes. This also introduces a parent abstract class `YarnSchedulerBackend` to encapsulate common logic to communicate with the Yarn `ApplicationMaster`. (2) The interface of requesting executors exposed to the `SparkContext` is the same, but the communication between the scheduler backend and the AM uses total number executors desired instead of an incremental number. This is discussed in #2746 and explained in the comments in the code.
I have tested this significantly on a stable Yarn cluster.
------------
A remaining task for this issue is to tone down the error messages emitted when an executor is removed.
Currently, `SparkContext` and its components react as if the executor has failed, resulting in many scary error messages and eventual timeouts. While it's not strictly necessary to fix this as of the first-cut implementation of this mechanism, it would be good to add logic to distinguish this case. I prefer to address this in a separate PR. I have filed a separate JIRA for this task at SPARK-4134.
Author: Andrew Or <andrew@databricks.com>
Author: Andrew Or <andrewor14@gmail.com>
Closes#2840 from andrewor14/yarn-scaling-mechanism and squashes the following commits:
485863e [Andrew Or] Minor log message changes
4920be8 [Andrew Or] Clarify that public API is only for Yarn mode for now
1c57804 [Andrew Or] Reword a few comments + other review comments
6321140 [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-scaling-mechanism
02836c0 [Andrew Or] Limit scope of synchronization
4e2ed7f [Andrew Or] Fix bug: keep track of removed executors properly
73ade46 [Andrew Or] Wording changes (minor)
2a7a6da [Andrew Or] Add `sc.killExecutor` as a shorthand (minor)
665f229 [Andrew Or] Mima excludes
79aa2df [Andrew Or] Simplify the request interface by asking for a total
04f625b [Andrew Or] Fix race condition that causes over-allocation of executors
f4783f8 [Andrew Or] Change the semantics of requesting executors
005a124 [Andrew Or] Fix tests
4628b16 [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-scaling-mechanism
db4a679 [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-scaling-mechanism
572f5c5 [Andrew Or] Unused import (minor)
f30261c [Andrew Or] Kill multiple executors rather than one at a time
de260d9 [Andrew Or] Simplify by skipping useless null check
9c52542 [Andrew Or] Simplify by skipping the TaskSchedulerImpl
97dd1a8 [Andrew Or] Merge branch 'master' of github.com:apache/spark into yarn-scaling-mechanism
d987b3e [Andrew Or] Move addWebUIFilters to Yarn scheduler backend
7b76d0a [Andrew Or] Expose mechanism in SparkContext as developer API
47466cd [Andrew Or] Refactor common Yarn scheduler backend logic
c4dfaac [Andrew Or] Avoid thrashing when removing executors
53e8145 [Andrew Or] Start yarn actor early to listen for AM registration message
bbee669 [Andrew Or] Add mechanism in yarn client mode
This PR encapsulates #2330, which is itself a continuation of #2240. The first goal of this PR is to provide an alternate, simpler implementation of the ConnectionManager which is based on Netty.
In addition to this goal, however, we want to resolve [SPARK-3796](https://issues.apache.org/jira/browse/SPARK-3796), which calls for a standalone shuffle service which can be integrated into the YARN NodeManager, Standalone Worker, or on its own. This PR makes the first step in this direction by ensuring that the actual Netty service is as small as possible and extracted from Spark core. Given this, we should be able to construct this standalone jar which can be included in other JVMs without incurring significant dependency or runtime issues. The actual work to ensure that such a standalone shuffle service would work in Spark will be left for a future PR, however.
In order to minimize dependencies and allow for the service to be long-running (possibly much longer-running than Spark, and possibly having to support multiple version of Spark simultaneously), the entire service has been ported to Java, where we have full control over the binary compatibility of the components and do not depend on the Scala runtime or version.
These issues: have been addressed by folding in #2330:
SPARK-3453: Refactor Netty module to use BlockTransferService interface
SPARK-3018: Release all buffers upon task completion/failure
SPARK-3002: Create a connection pool and reuse clients across different threads
SPARK-3017: Integration tests and unit tests for connection failures
SPARK-3049: Make sure client doesn't block when server/connection has error(s)
SPARK-3502: SO_RCVBUF and SO_SNDBUF should be bootstrap childOption, not option
SPARK-3503: Disable thread local cache in PooledByteBufAllocator
TODO before mergeable:
- [x] Implement uploadBlock()
- [x] Unit tests for RPC side of code
- [x] Performance testing (see comments [here](https://github.com/apache/spark/pull/2753#issuecomment-59475022))
- [x] Turn OFF by default (currently on for unit testing)
Author: Reynold Xin <rxin@apache.org>
Author: Aaron Davidson <aaron@databricks.com>
Author: cocoatomo <cocoatomo77@gmail.com>
Author: Patrick Wendell <pwendell@gmail.com>
Author: Prashant Sharma <prashant.s@imaginea.com>
Author: Davies Liu <davies.liu@gmail.com>
Author: Anand Avati <avati@redhat.com>
Closes#2753 from aarondav/netty and squashes the following commits:
cadfd28 [Aaron Davidson] Turn netty off by default
d7be11b [Aaron Davidson] Turn netty on by default
4a204b8 [Aaron Davidson] Fail block fetches if client connection fails
2b0d1c0 [Aaron Davidson] 100ch
0c5bca2 [Aaron Davidson] Merge branch 'master' of https://github.com/apache/spark into netty
14e37f7 [Aaron Davidson] Address Reynold's comments
8dfcceb [Aaron Davidson] Merge branch 'master' of https://github.com/apache/spark into netty
322dfc1 [Aaron Davidson] Address Reynold's comments, including major rename
e5675a4 [Aaron Davidson] Fail outstanding RPCs as well
ccd4959 [Aaron Davidson] Don't throw exception if client immediately fails
9da0bc1 [Aaron Davidson] Add RPC unit tests
d236dfd [Aaron Davidson] Remove no-op serializer :)
7b7a26c [Aaron Davidson] Fix Nio compile issue
dd420fd [Aaron Davidson] Merge branch 'master' of https://github.com/apache/spark into netty-test
939f276 [Aaron Davidson] Attempt to make comm. bidirectional
aa58f67 [cocoatomo] [SPARK-3909][PySpark][Doc] A corrupted format in Sphinx documents and building warnings
8dc1ded [cocoatomo] [SPARK-3867][PySpark] ./python/run-tests failed when it run with Python 2.6 and unittest2 is not installed
5b5dbe6 [Prashant Sharma] [SPARK-2924] Required by scala 2.11, only one fun/ctor amongst overriden alternatives, can have default argument(s).
2c5d9dc [Patrick Wendell] HOTFIX: Fix build issue with Akka 2.3.4 upgrade.
020691e [Davies Liu] [SPARK-3886] [PySpark] use AutoBatchedSerializer by default
ae4083a [Anand Avati] [SPARK-2805] Upgrade Akka to 2.3.4
29c6dcf [Aaron Davidson] [SPARK-3453] Netty-based BlockTransferService, extracted from Spark core
f7e7568 [Reynold Xin] Fixed spark.shuffle.io.receiveBuffer setting.
5d98ce3 [Reynold Xin] Flip buffer.
f6c220d [Reynold Xin] Merge with latest master.
407e59a [Reynold Xin] Fix style violation.
a0518c7 [Reynold Xin] Implemented block uploads.
4b18db2 [Reynold Xin] Copy the buffer in fetchBlockSync.
bec4ea2 [Reynold Xin] Removed OIO and added num threads settings.
1bdd7ee [Reynold Xin] Fixed tests.
d68f328 [Reynold Xin] Logging close() in case close() fails.
f63fb4c [Reynold Xin] Add more debug message.
6afc435 [Reynold Xin] Added logging.
c066309 [Reynold Xin] Implement java.io.Closeable interface.
519d64d [Reynold Xin] Mark private package visibility and MimaExcludes.
f0a16e9 [Reynold Xin] Fixed test hanging.
14323a5 [Reynold Xin] Removed BlockManager.getLocalShuffleFromDisk.
b2f3281 [Reynold Xin] Added connection pooling.
d23ed7b [Reynold Xin] Incorporated feedback from Norman: - use same pool for boss and worker - remove ioratio - disable caching of byte buf allocator - childoption sendbuf/receivebuf - fire exception through pipeline
9e0cb87 [Reynold Xin] Fixed BlockClientHandlerSuite
5cd33d7 [Reynold Xin] Fixed style violation.
cb589ec [Reynold Xin] Added more test cases covering cleanup when fault happens in ShuffleBlockFetcherIteratorSuite
1be4e8e [Reynold Xin] Shorten NioManagedBuffer and NettyManagedBuffer class names.
108c9ed [Reynold Xin] Forgot to add TestSerializer to the commit list.
b5c8d1f [Reynold Xin] Fixed ShuffleBlockFetcherIteratorSuite.
064747b [Reynold Xin] Reference count buffers and clean them up properly.
2b44cf1 [Reynold Xin] Added more documentation.
1760d32 [Reynold Xin] Use Epoll.isAvailable in BlockServer as well.
165eab1 [Reynold Xin] [SPARK-3453] Refactor Netty module to use BlockTransferService.
Sorter uses generic-typed key for sorting. When data is large, it creates lots of key objects, which is not efficient. We should reuse the key in Sorter for memory efficiency. This change is part of the petabyte sort implementation from rxin .
The `Sorter` class was written in Java and marked package private. So it is only available to `org.apache.spark.util.collection`. I renamed it to `TimSort` and add a simple wrapper of it, still called `Sorter`, in Scala, which is `private[spark]`.
The benchmark code is updated, which now resets the array before each run. Here is the result on sorting primitive Int arrays of size 25 million using Sorter:
~~~
[info] - Sorter benchmark for key-value pairs !!! IGNORED !!!
Java Arrays.sort() on non-primitive int array: Took 13237 ms
Java Arrays.sort() on non-primitive int array: Took 13320 ms
Java Arrays.sort() on non-primitive int array: Took 15718 ms
Java Arrays.sort() on non-primitive int array: Took 13283 ms
Java Arrays.sort() on non-primitive int array: Took 13267 ms
Java Arrays.sort() on non-primitive int array: Took 15122 ms
Java Arrays.sort() on non-primitive int array: Took 15495 ms
Java Arrays.sort() on non-primitive int array: Took 14877 ms
Java Arrays.sort() on non-primitive int array: Took 16429 ms
Java Arrays.sort() on non-primitive int array: Took 14250 ms
Java Arrays.sort() on non-primitive int array: (13878 ms first try, 14499 ms average)
Java Arrays.sort() on primitive int array: Took 2683 ms
Java Arrays.sort() on primitive int array: Took 2683 ms
Java Arrays.sort() on primitive int array: Took 2701 ms
Java Arrays.sort() on primitive int array: Took 2746 ms
Java Arrays.sort() on primitive int array: Took 2685 ms
Java Arrays.sort() on primitive int array: Took 2735 ms
Java Arrays.sort() on primitive int array: Took 2669 ms
Java Arrays.sort() on primitive int array: Took 2693 ms
Java Arrays.sort() on primitive int array: Took 2680 ms
Java Arrays.sort() on primitive int array: Took 2642 ms
Java Arrays.sort() on primitive int array: (2948 ms first try, 2691 ms average)
Sorter without key reuse on primitive int array: Took 10732 ms
Sorter without key reuse on primitive int array: Took 12482 ms
Sorter without key reuse on primitive int array: Took 10718 ms
Sorter without key reuse on primitive int array: Took 12650 ms
Sorter without key reuse on primitive int array: Took 10747 ms
Sorter without key reuse on primitive int array: Took 10783 ms
Sorter without key reuse on primitive int array: Took 12721 ms
Sorter without key reuse on primitive int array: Took 10604 ms
Sorter without key reuse on primitive int array: Took 10622 ms
Sorter without key reuse on primitive int array: Took 11843 ms
Sorter without key reuse on primitive int array: (11089 ms first try, 11390 ms average)
Sorter with key reuse on primitive int array: Took 5141 ms
Sorter with key reuse on primitive int array: Took 5298 ms
Sorter with key reuse on primitive int array: Took 5066 ms
Sorter with key reuse on primitive int array: Took 5164 ms
Sorter with key reuse on primitive int array: Took 5203 ms
Sorter with key reuse on primitive int array: Took 5274 ms
Sorter with key reuse on primitive int array: Took 5186 ms
Sorter with key reuse on primitive int array: Took 5159 ms
Sorter with key reuse on primitive int array: Took 5164 ms
Sorter with key reuse on primitive int array: Took 5078 ms
Sorter with key reuse on primitive int array: (5311 ms first try, 5173 ms average)
~~~
So with key reuse, it is faster and less likely to trigger GC.
Author: Xiangrui Meng <meng@databricks.com>
Author: Reynold Xin <rxin@apache.org>
Closes#2937 from mengxr/SPARK-4084 and squashes the following commits:
d73c3d0 [Xiangrui Meng] address comments
0b7b682 [Xiangrui Meng] fix mima
a72f53c [Xiangrui Meng] update timeIt
38ba50c [Xiangrui Meng] update timeIt
720f731 [Xiangrui Meng] add doc about JIT specialization
78f2879 [Xiangrui Meng] update tests
7de2efd [Xiangrui Meng] update the Sorter benchmark code to be correct
8626356 [Xiangrui Meng] add prepare to timeIt and update testsin SorterSuite
5f0d530 [Xiangrui Meng] update method modifiers of SortDataFormat
6ffbe66 [Xiangrui Meng] rename Sorter to TimSort and add a Scala wrapper that is private[spark]
b00db4d [Xiangrui Meng] doc and tests
cf94e8a [Xiangrui Meng] renaming
464ddce [Reynold Xin] cherry-pick rxin's commit
A global UTF8 constant is very helpful to handle encoding problems when converting between String and bytes. There are several solutions here:
1. Add `val UTF_8 = Charset.forName("UTF-8")` to Utils.scala
2. java.nio.charset.StandardCharsets.UTF_8 (require JDK7)
3. io.netty.util.CharsetUtil.UTF_8
4. com.google.common.base.Charsets.UTF_8
5. org.apache.commons.lang.CharEncoding.UTF_8
6. org.apache.commons.lang3.CharEncoding.UTF_8
IMO, I prefer option 1) because people can find it easily.
This is a PR for option 1) and only fixes Spark Core.
Author: zsxwing <zsxwing@gmail.com>
Closes#2781 from zsxwing/SPARK-3922 and squashes the following commits:
f974edd [zsxwing] Merge branch 'master' into SPARK-3922
2d27423 [zsxwing] Refactor spark-core to use Refactor spark-core to use Utils.UTF_8
This avoids reading torrent broadcast variables when they are referenced in the closure but not used in the closure. This is done by using a `lazy val` to read broadcast blocks
cc rxin JoshRosen for review
Author: Shivaram Venkataraman <shivaram@cs.berkeley.edu>
Closes#2871 from shivaram/broadcast-read-value and squashes the following commits:
1456d65 [Shivaram Venkataraman] Use getUsedTimeMs and remove readObject
d6c5ee9 [Shivaram Venkataraman] Use laxy val to implement readBroadcastBlock
0b34df7 [Shivaram Venkataraman] Merge branch 'master' of https://github.com/apache/spark into broadcast-read-value
9cec507 [Shivaram Venkataraman] Test if broadcast variables are read lazily
768b40b [Shivaram Venkataraman] Merge branch 'master' of https://github.com/apache/spark into broadcast-read-value
8792ed8 [Shivaram Venkataraman] Make torrent broadcast read blocks on use. This avoids reading broadcast variables when they are referenced in the closure but not used by the code.
`read()` may return fewer bytes than requested; when this occurred, the old code would silently return less data than requested, which might cause stream corruption errors. `skip()` faces similar issues, too.
This patch fixes several cases where we mis-handle these methods' return values.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2969 from JoshRosen/file-channel-read-fix and squashes the following commits:
e724a9f [Josh Rosen] Fix similar issue of not checking skip() return value.
cbc03ce [Josh Rosen] Update the other log message, too.
01e6015 [Josh Rosen] file.getName -> file.getAbsolutePath
d961d95 [Josh Rosen] Fix another issue in FileServerSuite.
b9265d2 [Josh Rosen] Fix a similar (minor) issue in TestUtils.
cd9d76f [Josh Rosen] Fix a similar error in Tachyon:
3db0008 [Josh Rosen] Fix a similar read() error in Utils.offsetBytes().
db985ed [Josh Rosen] Fix unsafe usage of FileChannel.read():
cc @rxin
Author: GuoQiang Li <witgo@qq.com>
Closes#2929 from witgo/SPARK-4064 and squashes the following commits:
20110f2 [GuoQiang Li] Modify the exception msg
3425225 [GuoQiang Li] review commits
2b07e49 [GuoQiang Li] If we create a lot of big broadcast variables, Spark may hang
This change replaces usages of colt with commons-math3 equivalents, and makes some minor necessary adjustments to related code and tests to match.
Author: Sean Owen <sowen@cloudera.com>
Closes#2928 from srowen/SPARK-4022 and squashes the following commits:
61a232f [Sean Owen] Fix failure due to different sampling in JavaAPISuite.sample()
16d66b8 [Sean Owen] Simplify seeding with call to reseedRandomGenerator
a1a78e0 [Sean Owen] Use Well19937c
31c7641 [Sean Owen] Fix Python Poisson test by choosing a different seed; about 88% of seeds should work but 1 didn't, it seems
5c9c67f [Sean Owen] Additional test fixes from review
d8f88e0 [Sean Owen] Replace colt with commons-math3. Some tests do not pass yet.
The patch takes advantage an API provided in Hadoop 2.5 that allows getting accurate data on Hadoop FileSystem bytes read. It eliminates the old method, which naively accepts the split size as the input bytes. An impact of this change will be that input metrics go away when using against Hadoop versions earlier thatn 2.5. I can add this back in, but my opinion is that no metrics are better than inaccurate metrics.
This is difficult to write a test for because we don't usually build against a version of Hadoop that contains the function we need. I've tested it manually on a pseudo-distributed cluster.
Author: Sandy Ryza <sandy@cloudera.com>
Closes#2087 from sryza/sandy-spark-2621 and squashes the following commits:
23010b8 [Sandy Ryza] Missing style fixes
74fc9bb [Sandy Ryza] Make getFSBytesReadOnThreadCallback private
1ab662d [Sandy Ryza] Clear things up a bit
984631f [Sandy Ryza] Switch from pull to push model and add test
7ef7b22 [Sandy Ryza] Add missing curly braces
219abc9 [Sandy Ryza] Fall back to split size
90dbc14 [Sandy Ryza] SPARK-2621. Update task InputMetrics incrementally
This change makes the destroy function public for broadcast variables. Motivation for the change is described in https://issues.apache.org/jira/browse/SPARK-4030.
This patch also logs where destroy was called from if a broadcast variable is used after destruction.
Author: Shivaram Venkataraman <shivaram@cs.berkeley.edu>
Closes#2922 from shivaram/broadcast-destroy and squashes the following commits:
a11abab [Shivaram Venkataraman] Fix scala style in Utils.scala
bed9c9d [Shivaram Venkataraman] Make destroy blocking by default
e80c1ab [Shivaram Venkataraman] Make destroy public for broadcast variables Also log where destroy was called from if a broadcast variable is used after destruction.
The shutdown hook of `DiskBlockManager` would remove localDirs. So do not need to register them with `Utils.registerShutdownDeleteDir`. It causes duplicate removal of these local dirs and corresponding exceptions.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#2826 from viirya/fix_duplicate_localdir_remove and squashes the following commits:
051d4b5 [Liang-Chi Hsieh] check dir existing and return empty List as default.
2b91a9c [Liang-Chi Hsieh] remove duplicate removal of local dirs.
This patch adds Selenium tests for Spark's web UI. To avoid adding extra
dependencies to the test environment, the tests use Selenium's HtmlUnitDriver,
which is pure-Java, instead of, say, ChromeDriver.
I added new tests to try to reproduce a few UI bugs reported on JIRA, namely
SPARK-3021, SPARK-2105, and SPARK-2527. I wasn't able to reproduce these bugs;
I suspect that the older ones might have been fixed by other patches.
In order to use HtmlUnitDriver, I added an explicit dependency on the
org.apache.httpcomponents version of httpclient in order to prevent jets3t's
older version from taking precedence on the classpath.
I also upgraded ScalaTest to 2.2.1.
Author: Josh Rosen <joshrosen@apache.org>
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2474 from JoshRosen/webui-selenium-tests and squashes the following commits:
fcc9e83 [Josh Rosen] scalautils -> scalactic package rename
510e54a [Josh Rosen] [SPARK-3616] Add basic Selenium tests to WebUISuite.
This follows https://github.com/apache/spark/pull/2893 , but does not completely fix SPARK-3359 either. This fixes minor scaladoc/javadoc issues that Javadoc 8 will treat as errors.
Author: Sean Owen <sowen@cloudera.com>
Closes#2909 from srowen/SPARK-3359 and squashes the following commits:
f62c347 [Sean Owen] Fix some javadoc issues that javadoc 8 considers errors. This is not all of the errors turned up when javadoc 8 runs on output of genjavadoc.
In tests, we may want to have BlockManagers of size < 1MB (spark.storage.unrollMemoryThreshold). However, these BlockManagers are useless because we can't unroll anything in them ever. At the very least we need to log a warning.
tdas
Author: Andrew Or <andrew@databricks.com>
Closes#2917 from andrewor14/unroll-safely-logging and squashes the following commits:
38947e3 [Andrew Or] Warn against starting a block manager that's too small
fd621b4 [Andrew Or] Warn against failure to reserve initial memory threshold
This pull request is a first step towards the implementation of a stable, pull-based progress / status API for Spark (see [SPARK-2321](https://issues.apache.org/jira/browse/SPARK-2321)). For now, I'd like to discuss the basic implementation, API names, and overall interface design. Once we arrive at a good design, I'll go back and add additional methods to expose more information via these API.
#### Design goals:
- Pull-based API
- Usable from Java / Scala / Python (eventually, likely with a wrapper)
- Can be extended to expose more information without introducing binary incompatibilities.
- Returns immutable objects.
- Don't leak any implementation details, preserving our freedom to change the implementation.
#### Implementation:
- Add public methods (`getJobInfo`, `getStageInfo`) to SparkContext to allow status / progress information to be retrieved.
- Add public interfaces (`SparkJobInfo`, `SparkStageInfo`) for our API return values. These interfaces consist entirely of Java-style getter methods. The interfaces are currently implemented in Java. I decided to explicitly separate the interface from its implementation (`SparkJobInfoImpl`, `SparkStageInfoImpl`) in order to prevent users from constructing these responses themselves.
-Allow an existing JobProgressListener to be used when constructing a live SparkUI. This allows us to re-use this listeners in the implementation of this status API. There are a few reasons why this listener re-use makes sense:
- The status API and web UI are guaranteed to show consistent information.
- These listeners are already well-tested.
- The same garbage-collection / information retention configurations can apply to both this API and the web UI.
- Extend JobProgressListener to maintain `jobId -> Job` and `stageId -> Stage` mappings.
The progress API methods are implemented in a separate trait that's mixed into SparkContext. This helps to avoid SparkContext.scala from becoming larger and more difficult to read.
Author: Josh Rosen <joshrosen@databricks.com>
Author: Josh Rosen <joshrosen@apache.org>
Closes#2696 from JoshRosen/progress-reporting-api and squashes the following commits:
e6aa78d [Josh Rosen] Add tests.
b585c16 [Josh Rosen] Accept SparkListenerBus instead of more specific subclasses.
c96402d [Josh Rosen] Address review comments.
2707f98 [Josh Rosen] Expose current stage attempt id
c28ba76 [Josh Rosen] Update demo code:
646ff1d [Josh Rosen] Document spark.ui.retainedJobs.
7f47d6d [Josh Rosen] Clean up SparkUI constructors, per Andrew's feedback.
b77b3d8 [Josh Rosen] Merge remote-tracking branch 'origin/master' into progress-reporting-api
787444c [Josh Rosen] Move status API methods into trait that can be mixed into SparkContext.
f9a9a00 [Josh Rosen] More review comments:
3dc79af [Josh Rosen] Remove creation of unused listeners in SparkContext.
249ca16 [Josh Rosen] Address several review comments:
da5648e [Josh Rosen] Add example of basic progress reporting in Java.
7319ffd [Josh Rosen] Add getJobIdsForGroup() and num*Tasks() methods.
cc568e5 [Josh Rosen] Add note explaining that interfaces should not be implemented outside of Spark.
6e840d4 [Josh Rosen] Remove getter-style names and "consistent snapshot" semantics:
08cbec9 [Josh Rosen] Begin to sketch the interfaces for a stable, public status API.
ac2d13a [Josh Rosen] Add jobId->stage, stageId->stage mappings in JobProgressListener
24de263 [Josh Rosen] Create UI listeners in SparkContext instead of in Tabs:
If classes implementing Serializable or Externalizable interfaces throw
exceptions other than IOException or ClassNotFoundException from their
(de)serialization methods, then this results in an unhelpful
"IOException: unexpected exception type" rather than the actual exception that
produced the (de)serialization error.
This patch fixes this by adding a utility method that re-wraps any uncaught
exceptions in IOException (unless they are already instances of IOException).
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2932 from JoshRosen/SPARK-4080 and squashes the following commits:
cd3a9be [Josh Rosen] [SPARK-4080] Only throw IOException from [write|read][Object|External].
https://issues.apache.org/jira/browse/SPARK-4067
currently , we call Utils.tryOrExit everywhere
AppClient
Executor
TaskSchedulerImpl
It makes the name of ExecutorUncaughtExceptionHandler unfit to the real case....
Author: Nan Zhu <nanzhu@Nans-MacBook-Pro.local>
Author: Nan Zhu <nanzhu@nans-mbp.home>
Closes#2913 from CodingCat/SPARK-4067 and squashes the following commits:
035ee3d [Nan Zhu] make RAT happy
e62e416 [Nan Zhu] add some general Exit code
a10b63f [Nan Zhu] refactor
In the existing code, each coarse-grained executor has two concurrently running actor systems. This causes many more error messages to be logged than necessary when the executor is lost or killed because we receive a disassociation event for each of these actor systems.
This is blocking #2840.
Author: Andrew Or <andrewor14@gmail.com>
Closes#2863 from andrewor14/executor-actor-system and squashes the following commits:
44ce2e0 [Andrew Or] Avoid starting two actor systems on each executor
In deploy.ClientArguments.isValidJarUrl, the url is checked as follows.
def isValidJarUrl(s: String): Boolean = s.matches("(.+):(.+)jar")
So, it allows like 'hdfs:file.jar' (no authority).
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2925 from sarutak/uri-syntax-check-improvement and squashes the following commits:
cf06173 [Kousuke Saruta] Improved URI syntax checking
If Spark lunched multiple executors in one host for one application, every executor would download it dependent files and jars (if not using local: url) independently. It maybe result in huge latency. In my case, it result in 20 seconds latency to download dependent jars(size about 17M) when I lunched 32 executors in every host(total 4 hosts).
This patch will cache downloaded files and jars for executors to reduce network throughput and download latency. In my case, the latency was reduced from 20 seconds to less than 1 second.
Author: Li Zhihui <zhihui.li@intel.com>
Author: li-zhihui <zhihui.li@intel.com>
Closes#1616 from li-zhihui/cachefiles and squashes the following commits:
36940df [Li Zhihui] Close cache for local mode
935fed6 [Li Zhihui] Clean code.
f9330d4 [Li Zhihui] Clean code again
7050d46 [Li Zhihui] Clean code
074a422 [Li Zhihui] Fix: deal with spark.files.overwrite
03ed3a8 [li-zhihui] rename cache file name as XXXXXXXXX_cache
2766055 [li-zhihui] Use url.hashCode + timestamp as cachedFileName
76a7b66 [Li Zhihui] Clean code & use applcation work directory as cache directory
3510eb0 [Li Zhihui] Keep fetchFile private
2ffd742 [Li Zhihui] add comment for FileLock
e0ebd48 [Li Zhihui] Try and finally lock.release
7fb7c0b [Li Zhihui] Release lock before copy files
6b997bf [Li Zhihui] Executors of same application in same host should only download files & jars once
After take(), maybe there are some garbage left in the socket, then next task assigned to this worker will hang because of corrupted data.
We should make sure the socket is clean before reuse it, write END_OF_STREAM at the end, and check it after read out all result from python.
Author: Davies Liu <davies.liu@gmail.com>
Author: Davies Liu <davies@databricks.com>
Closes#2838 from davies/fix_reuse and squashes the following commits:
8872914 [Davies Liu] fix tests
660875b [Davies Liu] fix bug while reuse worker after take()
This commit fixes a bug in MapStatus that could cause jobs to wrongly return
empty results if those jobs contained stages with more than 2000 partitions
where most of those partitions were empty.
For jobs with > 2000 partitions, MapStatus uses HighlyCompressedMapStatus,
which only stores the average size of blocks. If the average block size is
zero, then this will cause all blocks to be reported as empty, causing
BlockFetcherIterator to mistakenly skip them.
For example, this would return an empty result:
sc.makeRDD(0 until 10, 1000).repartition(2001).collect()
This can also lead to deserialization errors (e.g. Snappy decoding errors)
for jobs with > 2000 partitions where the average block size is non-zero but
there is at least one empty block. In this case, the BlockFetcher attempts to
fetch empty blocks and fails when trying to deserialize them.
The root problem here is that MapStatus has a (previously undocumented)
correctness property that was violated by HighlyCompressedMapStatus:
If a block is non-empty, then getSizeForBlock must be non-zero.
I fixed this by modifying HighlyCompressedMapStatus to store the average size
of _non-empty_ blocks and to use a compressed bitmap to track which blocks are
empty.
I also removed a test which was broken as originally written: it attempted
to check that HighlyCompressedMapStatus's size estimation error was < 10%,
but this was broken because HighlyCompressedMapStatus is only used for map
statuses with > 2000 partitions, but the test only created 50.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2866 from JoshRosen/spark-4019 and squashes the following commits:
fc8b490 [Josh Rosen] Roll back hashset change, which didn't improve performance.
5faa0a4 [Josh Rosen] Incorporate review feedback
c8b8cae [Josh Rosen] Two performance fixes:
3b892dd [Josh Rosen] Address Reynold's review comments
ba2e71c [Josh Rosen] Add missing newline
609407d [Josh Rosen] Use Roaring Bitmap to track non-empty blocks.
c23897a [Josh Rosen] Use sets when comparing collect() results
91276a3 [Josh Rosen] [SPARK-4019] Fix MapStatus compression bug that could lead to empty results.
...r without a remove in between. The cause for that is unknown, and assumed a temp network issue.
However, since the second register is with a BlockManagerId on a different port, blockManagerInfo.contains() returns false, while blockManagerIdByExecutor returns Some. This inconsistency is caught in a conditional statement that does System.exit(1), which is a huge robustness issue for us.
The fix - simply remove the old id from both maps during register when this happens. We are mimicking the behavior of expireDeadHosts(), by doing local cleanup of the maps before trying to add new ones.
Also - added some logging for register and unregister.
This is just like https://github.com/apache/spark/pull/2854 except it's on master
Author: Tal Sliwowicz <tal.s@taboola.com>
Closes#2886 from tsliwowicz/master-block-mgr-removal and squashes the following commits:
094d508 [Tal Sliwowicz] some more white space change undone
41a2217 [Tal Sliwowicz] some more whitspaces change undone
7bcfc3d [Tal Sliwowicz] whitspaces fix
df9d98f [Tal Sliwowicz] Code review comments fixed
f48bce9 [Tal Sliwowicz] In long running contexts, we encountered the situation of double register without a remove in between. The cause for that is unknown, and assumed a temp network issue.
This PR fixes SPARK-3426, an issue where sort-based shuffle crashes if the
`spark.shuffle.spill.compress` and `spark.shuffle.compress` settings have
different values.
The problem is that sort-based shuffle's read and write paths use different
settings for determining whether to apply compression. ExternalSorter writes
runs to files using `TempBlockId` ids, which causes
`spark.shuffle.spill.compress` to be used for enabling compression, but these
spilled files end up being shuffled over the network and read as shuffle files
using `ShuffleBlockId` by BlockStoreShuffleFetcher, which causes
`spark.shuffle.compress` to be used for enabling decompression. As a result,
this leads to errors when these settings disagree.
Based on the discussions in #2247 and #2178, it sounds like we don't want to
remove the `spark.shuffle.spill.compress` setting. Therefore, I've tried to
come up with a fix where `spark.shuffle.spill.compress` is used to compress
data that's read and written locally and `spark.shuffle.compress` is used to
compress any data that will be fetched / read as shuffle blocks.
To do this, I split `TempBlockId` into two new id types, `TempLocalBlockId` and
`TempShuffleBlockId`, which map to `spark.shuffle.spill.compress` and
`spark.shuffle.compress`, respectively. ExternalAppendOnlyMap also used temp
blocks for spilling data. It looks like ExternalSorter was designed to be
a generic sorter but its configuration already happens to be tied to sort-based
shuffle, so I think it's fine if we use `spark.shuffle.compress` to compress
its spills; we can move the compression configuration to the constructor in
a later commit if we find that ExternalSorter is being used in other contexts
where we want different configuration options to control compression. To
summarize:
**Before:**
| | ExternalAppendOnlyMap | ExternalSorter |
|-------|------------------------------|------------------------------|
| Read | spark.shuffle.spill.compress | spark.shuffle.compress |
| Write | spark.shuffle.spill.compress | spark.shuffle.spill.compress |
**After:**
| | ExternalAppendOnlyMap | ExternalSorter |
|-------|------------------------------|------------------------|
| Read | spark.shuffle.spill.compress | spark.shuffle.compress |
| Write | spark.shuffle.spill.compress | spark.shuffle.compress |
Thanks to andrewor14 for debugging this with me!
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2890 from JoshRosen/SPARK-3426 and squashes the following commits:
1921cf6 [Josh Rosen] Minor edit for clarity.
c8dd8f2 [Josh Rosen] Add comment explaining use of createTempShuffleBlock().
2c687b9 [Josh Rosen] Fix SPARK-3426.
91e7e40 [Josh Rosen] Combine tests into single test of all combinations
76ca65e [Josh Rosen] Add regression test for SPARK-3426.
runningLocally is deprecated now
Author: CrazyJvm <crazyjvm@gmail.com>
Closes#2879 from CrazyJvm/runningLocally and squashes the following commits:
bec0b3e [CrazyJvm] use isRunningLocally rather than runningLocally
Author: Sandy Ryza <sandy@cloudera.com>
Closes#789 from sryza/sandy-spark-1813 and squashes the following commits:
48b05e9 [Sandy Ryza] Simplify
b824932 [Sandy Ryza] Allow both spark.kryo.classesToRegister and spark.kryo.registrator at the same time
6a15bb7 [Sandy Ryza] Small fix
a2278c0 [Sandy Ryza] Respond to review comments
6ef592e [Sandy Ryza] SPARK-1813. Add a utility to SparkConf that makes using Kryo really easy
See [JIRA](https://issues.apache.org/jira/browse/SPARK-3994) for more information. Also adds
a note which warns against using these methods.
Author: Aaron Davidson <aaron@databricks.com>
Closes#2839 from aarondav/countByKey and squashes the following commits:
d6fdb2a [Aaron Davidson] Respond to comments
e1f06d3 [Aaron Davidson] [SPARK-3994] Use standard Aggregator code path for countByKey and countByValue
If an executor fails without being scheduled to run any tasks, then `DAGScheduler` won't notify `BlockManagerMasterActor` that the associated block manager should be removed. Instead, the associated block manager will be expired only after a few rounds of heartbeat timeouts. In terms of removal treatment, there should really be no distinction between executors that have been scheduled tasks and those that have not.
The fix, then, is to add all known executors to `TaskSchedulerImpl`'s `activeExecutorIds` whether or not it has been scheduled a task. In fact, the existing comment above `activeExecutorIds` is
```
// Which executor IDs we have executors on
val activeExecutorIds = new HashSet[String]
```
not "Which executors have been scheduled tasks thus far."
Author: Andrew Or <andrewor14@gmail.com>
Closes#2865 from andrewor14/active-executors and squashes the following commits:
ff3172b [Andrew Or] Add all known executors to `activeExecutorIds`
Just found a typo. Should not use "%f" for Long.
Author: zsxwing <zsxwing@gmail.com>
Closes#2875 from zsxwing/SPARK-4035 and squashes the following commits:
ce347e2 [zsxwing] Fix a wrong format specifier
This PR makes several changes to TorrentBroadcast in order to make
it easier to reason about, which should help when debugging SPARK-3958.
The key changes:
- Remove all state from the global TorrentBroadcast object. This state
consisted mainly of configuration options, like the block size and
compression codec, and was read by the blockify / unblockify methods.
Unfortunately, the use of `lazy val` for `BLOCK_SIZE` meant that the block
size was always determined by the first SparkConf that TorrentBroadast was
initialized with; as a result, unit tests could not properly test
TorrentBroadcast with different block sizes.
Instead, blockifyObject and unBlockifyObject now accept compression codecs
and blockSizes as arguments. These arguments are supplied at the call sites
inside of TorrentBroadcast instances. Each TorrentBroadcast instance
determines these values from SparkEnv's SparkConf. I was careful to ensure
that we do not accidentally serialize CompressionCodec or SparkConf objects
as part of the TorrentBroadcast object.
- Remove special-case handling of local-mode in TorrentBroadcast. I don't
think that broadcast implementations should know about whether we're running
in local mode. If we want to optimize the performance of broadcast in local
mode, then we should detect this at a higher level and use a dummy
LocalBroadcastFactory implementation instead.
Removing this code fixes a subtle error condition: in the old local mode
code, a failure to find the broadcast in the local BlockManager would lead
to an attempt to deblockify zero blocks, which could lead to confusing
deserialization or decompression errors when we attempted to decompress
an empty byte array. This should never have happened, though: a failure to
find the block in local mode is evidence of some other error. The changes
here will make it easier to debug those errors if they ever happen.
- Add a check that throws an exception when attempting to deblockify an
empty array.
- Use ScalaCheck to add a test to check that TorrentBroadcast's
blockifyObject and unBlockifyObject methods are inverses.
- Misc. cleanup and logging improvements.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2844 from JoshRosen/torrentbroadcast-bugfix and squashes the following commits:
1e8268d [Josh Rosen] Address Reynold's review comments
2a9fdfd [Josh Rosen] Address Reynold's review comments.
c3b08f9 [Josh Rosen] Update TorrentBroadcast tests to reflect removal of special local-mode optimizations.
5c22782 [Josh Rosen] Store broadcast variable's value in the driver.
33fc754 [Josh Rosen] Change blockify/unblockifyObject to accept serializer as argument.
618a872 [Josh Rosen] [SPARK-3958] TorrentBroadcast cleanup / debugging improvements.
Before, if the master node is killed and restarted, the worker nodes
would not attempt to reconnect to the Master. Therefore, when the Master
node was restarted, the worker nodes needed to be restarted as well.
Now, when the Master node is disconnected, the worker nodes will
continuously ping the master node in attempts to reconnect to it. Once
the master node restarts, it will detect one of the registration
requests from its former workers. The result is that the cluster
re-enters a healthy state.
In addition, when the master does not receive a heartbeat from the
worker, the worker was removed; however, when the worker sent a
heartbeat to the master, the master used to ignore the heartbeat. Now,
a master that receives a heartbeat from a worker that had been
disconnected will request the worker to re-attempt the registration
process, at which point the worker will send a RegisterWorker request
and be re-connected accordingly.
Re-connection attempts per worker are submitted every N seconds, where N
is configured by the property spark.worker.reconnect.interval - this has
a default of 60 seconds right now.
Author: mcheah <mcheah@palantir.com>
Closes#2828 from mccheah/reconnect-dead-workers and squashes the following commits:
83f8bc9 [mcheah] [SPARK-3736] More informative log message, and fixing some indentation.
fe0e02f [mcheah] [SPARK-3736] Moving reconnection logic to registerWithMaster().
94ddeca [mcheah] [SPARK-3736] Changing a log warning to a log info.
a698e35 [mcheah] [SPARK-3736] Addressing PR comment to make some defs private.
b9a3077 [mcheah] [SPARK-3736] Addressing PR comments related to reconnection.
2ad5ed5 [mcheah] [SPARK-3736] Cancel attempts to reconnect if the master changes.
b5b34af [mcheah] [SPARK-3736] Workers reconnect when disassociated from the master.
The problem caused by #1966
CC YanTangZhai andrewor14
Author: GuoQiang Li <witgo@qq.com>
Closes#2858 from witgo/SPARK-4010 and squashes the following commits:
9866fbf [GuoQiang Li] Spark UI returns 500 in yarn-client mode
Kernel 2.6.32 bug will lead to unexpected behavior of transferTo in copyStream, and this will corrupt the shuffle output file in sort-based shuffle, which will somehow introduce PARSING_ERROR(2), deserialization error or offset out of range. Here fix this by adding append flag, also add some position checking code. Details can be seen in [SPARK-3948](https://issues.apache.org/jira/browse/SPARK-3948).
Author: jerryshao <saisai.shao@intel.com>
Closes#2824 from jerryshao/SPARK-3948 and squashes the following commits:
be0533a [jerryshao] Address the comments
a82b184 [jerryshao] add configuration to control the NIO way of copying stream
e17ada2 [jerryshao] Fix kernel 2.6.32 bug led unexpected behavior of transferTo
This PR adds a Java API for AsyncRDDActions and promotes the API from `Experimental` to stable.
Author: Josh Rosen <joshrosen@apache.org>
Author: Josh Rosen <joshrosen@databricks.com>
Closes#2760 from JoshRosen/async-rdd-actions-in-java and squashes the following commits:
0d45fbc [Josh Rosen] Whitespace fix.
ad3ae53 [Josh Rosen] Merge remote-tracking branch 'origin/master' into async-rdd-actions-in-java
c0153a5 [Josh Rosen] Remove unused variable.
e8e2867 [Josh Rosen] Updates based on Marcelo's review feedback
7a1417f [Josh Rosen] Removed unnecessary java.util import.
6f8f6ac [Josh Rosen] Fix import ordering.
ff28e49 [Josh Rosen] Add MiMa excludes and fix a scalastyle error.
346e46e [Josh Rosen] [SPARK-3902] Stabilize AsyncRDDActions; add Java API.
This patch attempts to fix SPARK-2546 in `branch-1.0` and `branch-1.1`. The underlying problem is that thread-safety issues in Hadoop Configuration objects may cause Spark tasks to get stuck in infinite loops. The approach taken here is to clone a new copy of the JobConf for each task rather than sharing a single copy between tasks. Note that there are still Configuration thread-safety issues that may affect the driver, but these seem much less likely to occur in practice and will be more complex to fix (see discussion on the SPARK-2546 ticket).
This cloning is guarded by a new configuration option (`spark.hadoop.cloneConf`) and is disabled by default in order to avoid unexpected performance regressions for workloads that are unaffected by the Configuration thread-safety issues.
Author: Josh Rosen <joshrosen@apache.org>
Closes#2684 from JoshRosen/jobconf-fix-backport and squashes the following commits:
f14f259 [Josh Rosen] Add configuration option to control cloning of Hadoop JobConf.
b562451 [Josh Rosen] Remove unused jobConfCacheKey field.
dd25697 [Josh Rosen] [SPARK-2546] [1.0 / 1.1 backport] Clone JobConf for each task.
(cherry picked from commit 2cd40db2b3)
Signed-off-by: Josh Rosen <joshrosen@databricks.com>
Conflicts:
core/src/main/scala/org/apache/spark/rdd/HadoopRDD.scala
Make JavaPairRDD.collectAsMap result Serializable since Java Maps generally are
Author: Sean Owen <sowen@cloudera.com>
Closes#2805 from srowen/SPARK-3926 and squashes the following commits:
ecb78ee [Sean Owen] Fix conflict between java.io.Serializable and use of Scala's Serializable
f4717f9 [Sean Owen] Oops, fix compile problem
ae1b36f [Sean Owen] Expand to cover Maps returned from other Java API methods as well
51c26c2 [Sean Owen] Make JavaPairRDD.collectAsMap result Serializable since Java Maps generally are
There is a unused variable(count) in saveAsHadoopDataset in PairRDDFunctions.scala. The initial idea of this variable seems to count the number of records, so I am adding a log statement to log the number of records that has been written to the writer.
Author: likun <jacky.likun@huawei.com>
Author: jackylk <jacky.likun@huawei.com>
Closes#2791 from jackylk/SPARK-3935 and squashes the following commits:
a874047 [jackylk] removing the unused variable in PairRddFunctions.scala
3bf43c7 [likun] log the number of records has been written
Its hard to debug which broadcast variables refer to what in a big codebase. Printing call site information helps in debugging.
Author: Shivaram Venkataraman <shivaram@cs.berkeley.edu>
Closes#2829 from shivaram/spark-broadcast-print and squashes the following commits:
cd6dbdf [Shivaram Venkataraman] Print call site information for broadcasts
JobProgressPage could not show Fair Scheduler Pools section sometimes.
SparkContext starts webui and then postEnvironmentUpdate. Sometimes JobProgressPage is accessed between webui starting and postEnvironmentUpdate, then the lazy val isFairScheduler will be false. The Fair Scheduler Pools section will not display any more.
Author: yantangzhai <tyz0303@163.com>
Author: YanTangZhai <hakeemzhai@tencent.com>
Closes#1966 from YanTangZhai/SPARK-3067 and squashes the following commits:
d4323f8 [yantangzhai] update [SPARK-3067] JobProgressPage could not show Fair Scheduler Pools section sometimes
8a00106 [YanTangZhai] Merge pull request #6 from apache/master
b6391cc [yantangzhai] revert [SPARK-3067] JobProgressPage could not show Fair Scheduler Pools section sometimes
d2226cd [yantangzhai] [SPARK-3067] JobProgressPage could not show Fair Scheduler Pools section sometimes
cbcba66 [YanTangZhai] Merge pull request #3 from apache/master
aac7f7b [yantangzhai] [SPARK-3067] JobProgressPage could not show Fair Scheduler Pools section sometimes
cdef539 [YanTangZhai] Merge pull request #1 from apache/master
Sorry. I found that I forgot to add `afterExecute` for `handleConnectExecutor` in #2593.
Author: zsxwing <zsxwing@gmail.com>
Closes#2794 from zsxwing/SPARK-3741 and squashes the following commits:
a0bc4dd [zsxwing] Add afterExecute for handleConnectExecutor
In BlockManagermasterActor, _remainingMem would increase memSize for twice when updateBlockInfo if new storageLevel is invalid and old storageLevel is "useMemory". Also, _remainingMem should increase with original memory size instead of new memSize.
Author: Zhang, Liye <liye.zhang@intel.com>
Closes#2792 from liyezhang556520/spark-3941-remainMem and squashes the following commits:
3d487cc [Zhang, Liye] make the code concise
0380a32 [Zhang, Liye] [SPARK-3941][CORE] _remainingmem should not increase twice when updateBlockInfo
Something about the 2.3.4 upgrade seems to have made the issue manifest where all the services disconnect from each other after exactly 1000 seconds (which is the heartbeat interval). [This post](https://groups.google.com/forum/#!topic/akka-user/X3xzpTCbEFs) suggests that heartbeat pause should be greater than heartbeat interval, and increasing the pause from 600s to 6000s seems to have rectified the issue. My current cluster has now exceeded 1400s of uptime without failure!
I do not know why this fixed it, because the threshold we have set for the failure detector is the exponent of a timeout, and 300 is extremely large. Perhaps the default failure detector changed in 2.3.4 and now ignores threshold.
Author: Aaron Davidson <aaron@databricks.com>
Closes#2784 from aarondav/fix-timeout and squashes the following commits:
bd1151a [Aaron Davidson] Increase pause, don't decrease interval
9cb0372 [Aaron Davidson] [SPARK-3923] Decrease Akka heartbeat interval below heartbeat pause
This is a small number of clean-up changes on top of #2782. Closes#2782.
Author: Prashant Sharma <prashant.s@imaginea.com>
Author: Patrick Wendell <pwendell@gmail.com>
Closes#2803 from pwendell/pr-2782 and squashes the following commits:
56d5b7a [Patrick Wendell] Minor clean-up
44089ec [Patrick Wendell] Clean-up the TaskContext API.
ed551ce [Prashant Sharma] Fixed a typo
df261d0 [Prashant Sharma] Josh's suggestion
facf3b1 [Prashant Sharma] Fixed the mima issue.
7ecc2fe [Prashant Sharma] CR, Moved implementations to TaskContextImpl
bbd9e05 [Prashant Sharma] adding missed out files to git.
ef633f5 [Prashant Sharma] SPARK-3874, Provide stable TaskContext API
Customized pickler should be registered before unpickling, but in executor, there is no way to register the picklers before run the tasks.
So, we need to register the picklers in the tasks itself, duplicate the javaToPython() and pythonToJava() in MLlib, call SerDe.initialize() before pickling or unpickling.
Author: Davies Liu <davies.liu@gmail.com>
Closes#2830 from davies/fix_pickle and squashes the following commits:
0c85fb9 [Davies Liu] revert the privacy change
6b94e15 [Davies Liu] use JavaConverters instead of JavaConversions
0f02050 [Davies Liu] hotfix: Customized pickler does not work in cluster
Author: Shiti <ssaxena.ece@gmail.com>
Closes#2810 from Shiti/master and squashes the following commits:
051d82f [Shiti] setting the default value of uri scheme to "file" where matching "file" or None yields the same result
This is another implementation about #1256
cc andrewor14 vanzin
Author: GuoQiang Li <witgo@qq.com>
Closes#2379 from witgo/SPARK-2098-new and squashes the following commits:
4ef1cbd [GuoQiang Li] review commit
49ef70e [GuoQiang Li] Refactor getDefaultPropertiesFile
c45d20c [GuoQiang Li] All Spark processes should support spark-defaults.conf, config file
Author: shitis <ssaxena.ece@gmail.com>
Closes#2795 from Shiti/master and squashes the following commits:
46897d7 [shitis] Using Option Wrapper to convert String with value null to None
Validate the memory is greater than zero when set from the SPARK_WORKER_MEMORY environment variable or command line without a g or m label. Added unit tests. If memory is 0 an IllegalStateException is thrown. Updated unit tests to mock environment variables by subclassing SparkConf (tip provided by Josh Rosen). Updated WorkerArguments to use SparkConf.getenv instead of System.getenv for reading the SPARK_WORKER_MEMORY environment variable.
Author: Bill Bejeck <bbejeck@gmail.com>
Closes#2309 from bbejeck/spark-memory-worker and squashes the following commits:
51cf915 [Bill Bejeck] SPARK-3178 - Validate the memory is greater than zero when set from the SPARK_WORKER_MEMORY environment variable or command line without a g or m label. Added unit tests. If memory is 0 an IllegalStateException is thrown. Updated unit tests to mock environment variables by subclassing SparkConf (tip provided by Josh Rosen). Updated WorkerArguments to use SparkConf.getenv instead of System.getenv for reading the SPARK_WORKER_MEMORY environment variable.
The goal of this patch is to fix the swapped arguments in standalone mode, which was caused by 79e45c9323 (diff-79391110e9f26657e415aa169a004998R153).
More details can be found in the JIRA: [SPARK-3921](https://issues.apache.org/jira/browse/SPARK-3921)
Tested in Standalone mode, but not in Mesos.
Author: Aaron Davidson <aaron@databricks.com>
Closes#2779 from aarondav/fix-standalone and squashes the following commits:
725227a [Aaron Davidson] Fix ExecutorRunnerTest
9d703fe [Aaron Davidson] [SPARK-3921] Fix CoarseGrainedExecutorBackend's arguments for Standalone mode
In the comment (Line 1083), it says: "Otherwise, interpolate the number of partitions we need to try, but overestimate it by 50%."
`(1.5 * num * partsScanned / buf.size).toInt` is the guess of "num of total partitions needed". In every iteration, we should consider the increment `(1.5 * num * partsScanned / buf.size).toInt - partsScanned`
Existing implementation 'exponentially' grows `partsScanned ` ( roughly: `x_{n+1} >= (1.5 + 1) x_n`)
This could be a performance problem. (unless this is the intended behavior)
Author: yingjieMiao <yingjie@42go.com>
Closes#2648 from yingjieMiao/rdd_take and squashes the following commits:
d758218 [yingjieMiao] scala style fix
a8e74bb [yingjieMiao] python style fix
4b6e777 [yingjieMiao] infix operator style fix
4391d3b [yingjieMiao] typo fix.
692f4e6 [yingjieMiao] cap numPartsToTry
c4483dc [yingjieMiao] style fix
1d2c410 [yingjieMiao] also change in rdd.py and AsyncRDD
d31ff7e [yingjieMiao] handle the edge case after 1 iteration
a2aa36b [yingjieMiao] RDD take method: overestimate too much
Author: GuoQiang Li <witgo@qq.com>
Closes#2763 from witgo/SPARK-3905 and squashes the following commits:
17d7990 [GuoQiang Li] The keys for sorting the columns of Executor page ,Stage page Storage page are incorrect
val path = ... //path to seq file with BytesWritable as type of both key and value
val file = sc.sequenceFile[Array[Byte],Array[Byte]](path)
file.take(1)(0)._1
This prints incorrect content of byte array. Actual content starts with correct one and some "random" bytes and zeros are appended. BytesWritable has two methods:
getBytes() - return content of all internal array which is often longer then actual value stored. It usually contains the rest of previous longer values
copyBytes() - return just begining of internal array determined by internal length property
It looks like in implicit conversion between BytesWritable and Array[byte] getBytes is used instead of correct copyBytes.
dbtsai
Author: Jakub Dubovský <james64@inMail.sk>
Author: Dubovsky Jakub <dubovsky@avast.com>
Closes#2712 from james64/3121-bugfix and squashes the following commits:
f85d24c [Jakub Dubovský] Test name changed, comments added
1b20d51 [Jakub Dubovský] Import placed correctly
406e26c [Jakub Dubovský] Scala style fixed
f92ffa6 [Dubovsky Jakub] performance tuning
480f9cd [Dubovsky Jakub] Bug 3121 fixed
When reporting that a remote error occurred, the ConnectionManager should also log the stacktrace of the remote exception. This PR accomplishes this by sending the remote exception's stacktrace as the payload in the "negative ACK / error message."
Author: Josh Rosen <joshrosen@apache.org>
Closes#2741 from JoshRosen/propagate-cm-exceptions-to-sender and squashes the following commits:
b5366cc [Josh Rosen] Explicitly encode error messages using UTF-8.
cef18b3 [Josh Rosen] [SPARK-3887] Send stracktrace in ConnectionManager error messages.
...riden alternatives, can have default argument.
Author: Prashant Sharma <prashant.s@imaginea.com>
Closes#2750 from ScrapCodes/SPARK-2924/default-args-removed and squashes the following commits:
d9785c3 [Prashant Sharma] [SPARK-2924] Required by scala 2.11, only one function/ctor amongst overriden alternatives, can have default argument.
In general, individual shuffle blocks are frequently small, so mmapping them often creates a lot of waste. It may not be bad to mmap the larger ones, but it is pretty inconvenient to get configuration into ManagedBuffer, and besides it is unlikely to help all that much.
Author: Aaron Davidson <aaron@databricks.com>
Closes#2742 from aarondav/mmap and squashes the following commits:
a152065 [Aaron Davidson] Add other pathway back
52b6cd2 [Aaron Davidson] [SPARK-3889] Attempt to avoid SIGBUS by not mmapping files in ConnectionManager
This is a second rev of the Akka upgrade (earlier merged, but reverted). I made a slight modification which is that I also upgrade Hive to deal with a compatibility issue related to the protocol buffers library.
Author: Anand Avati <avati@redhat.com>
Author: Patrick Wendell <pwendell@gmail.com>
Closes#2752 from pwendell/akka-upgrade and squashes the following commits:
4c7ca3f [Patrick Wendell] Upgrading to new hive->protobuf version
57a2315 [Anand Avati] SPARK-1812: streaming - remove tests which depend on akka.actor.IO
2a551d3 [Anand Avati] SPARK-1812: core - upgrade to akka 2.3.4
I noticed a few issues with how temp directories are created and deleted:
*Minor*
* Guava's `Files.createTempDir()` plus `File.deleteOnExit()` is used in many tests to make a temp dir, but `Utils.createTempDir()` seems to be the standard Spark mechanism
* Call to `File.deleteOnExit()` could be pushed into `Utils.createTempDir()` as well, along with this replacement
* _I messed up the message in an exception in `Utils` in SPARK-3794; fixed here_
*Bit Less Minor*
* `Utils.deleteRecursively()` fails immediately if any `IOException` occurs, instead of trying to delete any remaining files and subdirectories. I've observed this leave temp dirs around. I suggest changing it to continue in the face of an exception and throw one of the possibly several exceptions that occur at the end.
* `Utils.createTempDir()` will add a JVM shutdown hook every time the method is called. Even if the subdir is the parent of another parent dir, since this check is inside the hook. However `Utils` manages a set of all dirs to delete on shutdown already, called `shutdownDeletePaths`. A single hook can be registered to delete all of these on exit. This is how Tachyon temp paths are cleaned up in `TachyonBlockManager`.
I noticed a few other things that might be changed but wanted to ask first:
* Shouldn't the set of dirs to delete be `File`, not just `String` paths?
* `Utils` manages the set of `TachyonFile` that have been registered for deletion, but the shutdown hook is managed in `TachyonBlockManager`. Should this logic not live together, and not in `Utils`? it's more specific to Tachyon, and looks a slight bit odd to import in such a generic place.
Author: Sean Owen <sowen@cloudera.com>
Closes#2670 from srowen/SPARK-3811 and squashes the following commits:
071ae60 [Sean Owen] Update per @vanzin's review
da0146d [Sean Owen] Make Utils.deleteRecursively try to delete all paths even when an exception occurs; use one shutdown hook instead of one per method call to delete temp dirs
3a0faa4 [Sean Owen] Standardize on Utils.createTempDir instead of Files.createTempDir
This pull request addresses a few issues related to PySpark's IPython support:
- Fix the remaining uses of the '-u' flag, which IPython doesn't support (see SPARK-3772).
- Change PYSPARK_PYTHON_OPTS to PYSPARK_DRIVER_PYTHON_OPTS, so that the old name is reserved in case we ever want to allow the worker Python options to be customized (this variable was introduced in #2554 and hasn't landed in a release yet, so this doesn't break any compatibility).
- Introduce a PYSPARK_DRIVER_PYTHON option that allows the driver to use `ipython` while the workers use a different Python version.
- Attempt to use Python 2.7 by default if PYSPARK_PYTHON is not specified.
- Retain the old semantics for IPYTHON=1 and IPYTHON_OPTS (to avoid breaking existing example programs).
There are more details in a block comment in `bin/pyspark`.
Author: Josh Rosen <joshrosen@apache.org>
Closes#2651 from JoshRosen/SPARK-3772 and squashes the following commits:
7b8eb86 [Josh Rosen] More changes to PySpark python executable configuration:
c4f5778 [Josh Rosen] [SPARK-3772] Allow ipython to be used by Pyspark workers; IPython fixes:
...re logs to avoid Executors swallowing errors
This PR made the following changes:
* Register a callback to `Connection` so that the error will be propagated properly.
* Add more logs so that the errors won't be swallowed by Executors.
* Use trySuccess/tryFailure because `Promise` doesn't allow to call success/failure more than once.
Author: zsxwing <zsxwing@gmail.com>
Closes#2593 from zsxwing/SPARK-3741 and squashes the following commits:
1d5aed5 [zsxwing] Fix naming
0b8a61c [zsxwing] Merge branch 'master' into SPARK-3741
764aec5 [zsxwing] [SPARK-3741] Make ConnectionManager propagate errors properly and add more logs to avoid Executors swallowing errors
Truncate appName in WebUI if it is too long.
Author: Xiangrui Meng <meng@databricks.com>
Closes#2707 from mengxr/truncate-app-name and squashes the following commits:
87834ce [Xiangrui Meng] move scala import below java
c7111dc [Xiangrui Meng] truncate appName in WebUI if it is too long
Upgrade to akka 2.3.4
Author: Anand Avati <avati@redhat.com>
Closes#1685 from avati/SPARK-1812-akka-2.3 and squashes the following commits:
57a2315 [Anand Avati] SPARK-1812: streaming - remove tests which depend on akka.actor.IO
2a551d3 [Anand Avati] SPARK-1812: core - upgrade to akka 2.3.4
There is a Spark logo on the header of HistoryPage.
We can have too many HistoryPages if we run 20+ applications. So I think, it's useful if the logo is as a link to the HistoryPage's page number 1.
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2690 from sarutak/SPARK-3829 and squashes the following commits:
908c109 [Kousuke Saruta] Removed extra space.
00bfbd7 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-3829
dd87480 [Kousuke Saruta] Made header Spark log image as a link to History Server's top page.
Now the Stage page only displays "Executor"(host) for tasks. However, there may be more than one Executors running in the same host. Currently, when some task is hung, I only know the host of the faulty executor. Therefore I have to check all executors in the host.
Adding "Executor ID" in the Tasks table. would be helpful to locate the faulty executor. Here is the new page:
Author: zsxwing <zsxwing@gmail.com>
Closes#2642 from zsxwing/SPARK-3777 and squashes the following commits:
37945af [zsxwing] Put Executor ID and Host into one cell
4bbe2c7 [zsxwing] [SPARK-3777] Display "Executor ID" for Tasks in Stage page
Before:
```
14/10/06 16:45:42 WARN CacheManager: Not enough space to cache partition rdd_0_2
in memory! Free memory is 481861527 bytes.
```
After:
```
14/10/07 11:08:24 WARN MemoryStore: Not enough space to cache rdd_2_0 in memory!
(computed 68.8 MB so far)
14/10/07 11:08:24 INFO MemoryStore: Memory use = 1088.0 B (blocks) + 445.1 MB
(scratch space shared across 8 thread(s)) = 445.1 MB. Storage limit = 459.5 MB.
```
Author: Andrew Or <andrewor14@gmail.com>
Closes#2688 from andrewor14/cache-log-message and squashes the following commits:
28e33d6 [Andrew Or] Shy away from "unrolling"
5638c49 [Andrew Or] Grammar
39a0c28 [Andrew Or] Log more detail when unrolling a block fails
The parent.getOrCompute() of PythonRDD is executed in a separated thread, it should release the memory reserved for shuffle and unrolling finally.
Author: Davies Liu <davies.liu@gmail.com>
Closes#2668 from davies/leak and squashes the following commits:
ae98be2 [Davies Liu] fix memory leak in PythonRDD
SparkEnv is cached in ThreadLocal object, so after stop and create a new SparkContext, old SparkEnv is still used by some threads, it will trigger many problems, for example, pyspark will have problem after restart SparkContext, because py4j use thread pool for RPC.
This patch will clear all the references after stop a SparkEnv.
cc mateiz tdas pwendell
Author: Davies Liu <davies.liu@gmail.com>
Closes#2624 from davies/env and squashes the following commits:
a69f30c [Davies Liu] deprecate getThreadLocal
ba77ca4 [Davies Liu] remove getThreadLocal(), update docs
ee62bb7 [Davies Liu] cleanup ThreadLocal of SparnENV
4d0ea8b [Davies Liu] clear reference of SparkEnv after stop
With Spark SQL we generate very long RDD names. These names are not properly rendered in the web UI.
This PR fixes the rendering issue.
[SPARK-3827] #comment Linking PR with JIRA
Author: Hossein <hossein@databricks.com>
Closes#2687 from falaki/sparkTableUI and squashes the following commits:
fd06409 [Hossein] Limit width of cell when RDD name is too long
AccumulableParam gave its generic parameters as 'R, T', whereas SparkContext labeled them 'T, R'.
Trivial, but really confusing.
I resolved this in favor of AccumulableParam, because it seemed to have some logic for its names. I also extended this minimal, but at least present, justification into the SparkContext comments.
Author: Nathan Kronenfeld <nkronenfeld@oculusinfo.com>
Closes#2637 from nkronenfeld/accumulators and squashes the following commits:
98d6b74 [Nathan Kronenfeld] Rectify gereneric parameter names between SparkContext and AccumulableParam
Remove references to Commons IO FileUtils and replace with pure Java version, which doesn't need to traverse the whole directory tree first.
I think this method could be refined further if it would be alright to rename it and its args and break it down into two methods. I'm starting with a simple recursive rendition.
Author: Sean Owen <sowen@cloudera.com>
Closes#2662 from srowen/SPARK-3794 and squashes the following commits:
4cd172f [Sean Owen] Remove references to Commons IO FileUtils and replace with pure Java version, which doesn't need to traverse the whole directory tree first
JIRA: https://issues.apache.org/jira/browse/SPARK-1656
Author: zsxwing <zsxwing@gmail.com>
Closes#577 from zsxwing/SPARK-1656 and squashes the following commits:
c431095 [zsxwing] Add a comment and fix the code style
2de96e5 [zsxwing] Make sure file will be deleted if exception happens
28b90dc [zsxwing] Update to follow the code style
4521d6e [zsxwing] Merge branch 'master' into SPARK-1656
afc3383 [zsxwing] Update to follow the code style
071fdd1 [zsxwing] SPARK-1656: Fix potential resource leaks
The MesosSchedulerBackend did not previously implement `killTask`,
resulting in an exception.
Author: Brenden Matthews <brenden@diddyinc.com>
Closes#2453 from brndnmtthws/implement-killtask and squashes the following commits:
23ddcdc [Brenden Matthews] [SPARK-3597][Mesos] Implement `killTask`.
First contribution to the project, so apologize for any significant errors.
This PR addresses [SPARK-1860]. The application directories are now cleaned up in a more conservative manner.
Previously, app-* directories were cleaned up if the directory's timestamp was older than a given time. However, the timestamp on a directory does not reflect the modification times of the files in that directory. Therefore, app-* directories were wiped out even if the files inside them were created recently and possibly being used by Executor tasks.
The solution is to change the cleanup logic to inspect all files within the app-* directory and only eliminate the app-* directory if all files in the directory are stale.
Author: mcheah <mcheah@palantir.com>
Closes#2609 from mccheah/worker-better-app-dir-cleanup and squashes the following commits:
87b5d03 [mcheah] [SPARK-1860] Using more string interpolation. Better error logging.
802473e [mcheah] [SPARK-1860] Cleaning up the logs generated when cleaning directories.
e0a1f2e [mcheah] [SPARK-1860] Fixing broken unit test.
77a9de0 [mcheah] [SPARK-1860] More conservative app directory cleanup.
This PR is another solution for #2250
I'm using codahale base MetricsSystem of Spark with JMX or Graphite, and I saw following 2 problems.
(1) When applications which have same spark.app.name run on cluster at the same time, some metrics names are mixed. For instance, if 2+ application is running on the cluster at the same time, each application emits the same named metric like "SparkPi.DAGScheduler.stage.failedStages" and Graphite cannot distinguish the metrics is for which application.
(2) When 2+ executors run on the same machine, JVM metrics of each executors are mixed. For instance, 2+ executors running on the same node can emit the same named metric "jvm.memory" and Graphite cannot distinguish the metrics is from which application.
And there is an similar issue. The directory for event logs is named using application name.
Application name is defined by user and the name can includes illegal character for path names.
Further more, the directory name consists of application name and System.currentTimeMillis even though each application has unique Application ID so if we run jobs which have same name, it's difficult to identify which directory is for which application.
Closes#2250Closes#1067
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2432 from sarutak/metrics-structure-improvement2 and squashes the following commits:
3288b2b [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
39169e4 [Kousuke Saruta] Fixed style
6570494 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
817e4f0 [Kousuke Saruta] Simplified MetricsSystem#buildRegistryName
67fa5eb [Kousuke Saruta] Unified MetricsSystem#registerSources and registerSinks in start
10be654 [Kousuke Saruta] Fixed style.
990c078 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
f0c7fba [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
59cc2cd [Kousuke Saruta] Modified SparkContextSchedulerCreationSuite
f9b6fb3 [Kousuke Saruta] Modified style.
2cf8a0f [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
389090d [Kousuke Saruta] Replaced taskScheduler.applicationId() with getApplicationId in SparkContext#postApplicationStart
ff45c89 [Kousuke Saruta] Added some test cases to MetricsSystemSuite
69c46a6 [Kousuke Saruta] Added warning logging logic to MetricsSystem#buildRegistryName
5cca0d2 [Kousuke Saruta] Added Javadoc comment to SparkContext#getApplicationId
16a9f01 [Kousuke Saruta] Added data types to be returned to some methods
6434b06 [Kousuke Saruta] Reverted changes related to ApplicationId
0413b90 [Kousuke Saruta] Deleted ApplicationId.java and ApplicationIdSuite.java
a42300c [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
0fc1b09 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
42bea55 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
248935d [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
f6af132 [Kousuke Saruta] Modified SchedulerBackend and TaskScheduler to return System.currentTimeMillis as an unique Application Id
1b8b53e [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
97cb85c [Kousuke Saruta] Modified confliction of MimExcludes
2cdd009 [Kousuke Saruta] Modified defailt implementation of applicationId
9aadb0b [Kousuke Saruta] Modified NetworkReceiverSuite to ensure "executor.start()" is finished in test "network receiver life cycle"
3011efc [Kousuke Saruta] Added ApplicationIdSuite.scala
d009c55 [Kousuke Saruta] Modified ApplicationId#equals to compare appIds
dfc83fd [Kousuke Saruta] Modified ApplicationId to implement Serializable
9ff4851 [Kousuke Saruta] Modified MimaExcludes.scala to ignore createTaskScheduler method in SparkContext
4567ffc [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
6a91b14 [Kousuke Saruta] Modified SparkContextSchedulerCreationSuite, ExecutorRunnerTest and EventLoggingListenerSuite
0325caf [Kousuke Saruta] Added ApplicationId.scala
0a2fc14 [Kousuke Saruta] Modified style
eabda80 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
0f890e6 [Kousuke Saruta] Modified SparkDeploySchedulerBackend and Master to pass baseLogDir instead f eventLogDir
bcf25bf [Kousuke Saruta] Modified directory name for EventLogs
28d4d93 [Kousuke Saruta] Modified SparkContext and EventLoggingListener so that the directory for EventLogs is named same for Application ID
203634e [Kousuke Saruta] Modified comment in SchedulerBackend#applicationId and TaskScheduler#applicationId
424fea4 [Kousuke Saruta] Modified the subclasses of TaskScheduler and SchedulerBackend so that they can return non-optional Unique Application ID
b311806 [Kousuke Saruta] Swapped last 2 arguments passed to CoarseGrainedExecutorBackend
8a2b6ec [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
086ee25 [Kousuke Saruta] Merge branch 'metrics-structure-improvement2' of github.com:sarutak/spark into metrics-structure-improvement2
e705386 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
36d2f7a [Kousuke Saruta] Added warning message for the situation we cannot get application id for the prefix for the name of metrics
eea6e19 [Kousuke Saruta] Modified CoarseGrainedMesosSchedulerBackend and MesosSchedulerBackend so that we can get Application ID
c229fbe [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
e719c39 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
4a93c7f [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement2
4776f9e [Kousuke Saruta] Modified MetricsSystemSuite.scala
efcb6e1 [Kousuke Saruta] Modified to add application id to metrics name
2ec848a [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
3ea7896 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
ead8966 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
08e627e [Kousuke Saruta] Revert "tmp"
7b67f5a [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
45bd33d [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
93e263a [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
848819c [Kousuke Saruta] Merge branch 'metrics-structure-improvement' of github.com:sarutak/spark into metrics-structure-improvement
912a637 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
e4a4593 [Kousuke Saruta] tmp
3e098d8 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
4603a39 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
fa7175b [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
15f88a3 [Kousuke Saruta] Modified MetricsSystem#buildRegistryName because conf.get does not return null when correspondin entry is absent
6f7dcd4 [Kousuke Saruta] Modified constructor of DAGSchedulerSource and BlockManagerSource because the instance of SparkContext is no longer used
6fc5560 [Kousuke Saruta] Modified sourceName of ExecutorSource, DAGSchedulerSource and BlockManagerSource
4e057c9 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into metrics-structure-improvement
85ffc02 [Kousuke Saruta] Revert "Modified sourceName of ExecutorSource, DAGSchedulerSource and BlockManagerSource"
868e326 [Kousuke Saruta] Modified MetricsSystem to set registry name with unique application-id and driver/executor-id
71609f5 [Kousuke Saruta] Modified sourceName of ExecutorSource, DAGSchedulerSource and BlockManagerSource
55debab [Kousuke Saruta] Modified SparkContext and Executor to set spark.executor.id to identifiers
4180993 [Kousuke Saruta] Modified SparkContext to retain spark.unique.app.name property in SparkConf
The existing code only considered one of the RMs when running in
Yarn HA mode, so it was possible to get errors if the active RM
was not registered in the filter.
The change makes use of a new API added to Yarn that returns all
proxy addresses, and falls back to the old behavior if the API
is not present. While there, I also made a change to look for the
scheme (http or https) being used by Yarn when building the proxy
URIs.
Since, in the case of multiple RMs, Yarn uses commas as a separator,
it was not possible anymore to use spark.filter.params to propagate
this information (which used commas to delimit different config params).
Instead, I added a new param (spark.filter.jsonParams) which expects
a JSON string containing a map with the config data. I chose not to
add it to the documentation at this point since I don't believe users
will use it directly.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#2469 from vanzin/SPARK-3606 and squashes the following commits:
aeb458a [Marcelo Vanzin] Undelete needed import.
65e400d [Marcelo Vanzin] Remove unused import.
d121883 [Marcelo Vanzin] Use separate config for each param instead of json.
04bc156 [Marcelo Vanzin] Review feedback.
4d4d6b9 [Marcelo Vanzin] [SPARK-3606] [yarn] Correctly configure AmIpFilter for Yarn HA.
Author: Brenden Matthews <brenden@diddyinc.com>
Closes#2401 from brndnmtthws/master and squashes the following commits:
4abaa5d [Brenden Matthews] [SPARK-3535][Mesos] Fix resource handling.
Update of PR #997.
With this PR, setting SPARK_CONF_DIR overrides SPARK_HOME/conf (not only spark-defaults.conf and spark-env).
Author: EugenCepoi <cepoi.eugen@gmail.com>
Closes#2481 from EugenCepoi/SPARK-2058 and squashes the following commits:
0bb32c2 [EugenCepoi] use orElse orNull and fixing trailing percent in compute-classpath.cmd
77f35d7 [EugenCepoi] SPARK-2058: Overriding SPARK_HOME/conf with SPARK_CONF_DIR
SparkSubmitDriverBootstrapper.scala now returns the exit code of the driver process, instead of always returning 0.
Author: Eric Eijkelenboom <ee@userreport.com>
Closes#2628 from ericeijkelenboom/master and squashes the following commits:
cc4a571 [Eric Eijkelenboom] Return the exit code of the driver process
pwendell, ```tryPort``` is not compatible with old code in last PR, this is to fix it.
And after discuss with srowen renamed the title to "avoid trying privileged port when request a non-privileged port". Plz refer to the discuss for detail.
Author: scwf <wangfei1@huawei.com>
Closes#2623 from scwf/1-1024 and squashes the following commits:
10a4437 [scwf] add comment
de3fd17 [scwf] do not try privileged port when request a non-privileged port
42cb0fa [scwf] make tryPort compatible with old code
cb8cc76 [scwf] do not use port 1 - 1024
If you turn authentication on and you are using a lot of executors. There is a chance that all the of the threads in the handleMessageExecutor could be waiting to send a message because they are blocked waiting on authentication to happen. This can cause a temporary deadlock until the connection times out.
To fix it, I got rid of the wait/notify and use a single outbox but only send security messages from it until authentication has completed.
Author: Thomas Graves <tgraves@apache.org>
Closes#2484 from tgravescs/cm_threads_auth and squashes the following commits:
a0a961d [Thomas Graves] give it a type
b6bc80b [Thomas Graves] Rework comments
d6d4175 [Thomas Graves] update from comments
081b765 [Thomas Graves] cleanup
4d7f8f5 [Thomas Graves] Change to not use wait/notify while waiting for authentication
If a block manager (say, A) wants to replicate a block and the node chosen for replication (say, B) is dead, then the attempt to send the block to B fails. However, this continues to fail indefinitely. Even if the driver learns about the demise of the B, A continues to try replicating to B and failing miserably.
The reason behind this bug is that A initially fetches a list of peers from the driver (when B was active), but never updates it after B is dead. This affects Spark Streaming as its receiver uses block replication.
The solution in this patch adds the following.
- Changed BlockManagerMaster to return all the peers of a block manager, rather than the requested number. It also filters out driver BlockManager.
- Refactored BlockManager's replication code to handle peer caching correctly.
+ The peer for replication is randomly selected. This is different from past behavior where for a node A, a node B was deterministically chosen for the lifetime of the application.
+ If replication fails to one node, the peers are refetched.
+ The peer cached has a TTL of 1 second to enable discovery of new peers and using them for replication.
- Refactored use of \<driver\> in BlockManager into a new method `BlockManagerId.isDriver`
- Added replication unit tests (replication was not tested till now, duh!)
This should not make a difference in performance of Spark workloads where replication is not used.
@andrewor14 @JoshRosen
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#2366 from tdas/replication-fix and squashes the following commits:
9690f57 [Tathagata Das] Moved replication tests to a new BlockManagerReplicationSuite.
0661773 [Tathagata Das] Minor changes based on PR comments.
a55a65c [Tathagata Das] Added a unit test to test replication behavior.
012afa3 [Tathagata Das] Bug fix
89f91a0 [Tathagata Das] Minor change.
68e2c72 [Tathagata Das] Made replication peer selection logic more efficient.
08afaa9 [Tathagata Das] Made peer selection for replication deterministic to block id
3821ab9 [Tathagata Das] Fixes based on PR comments.
08e5646 [Tathagata Das] More minor changes.
d402506 [Tathagata Das] Fixed imports.
4a20531 [Tathagata Das] Filtered driver block manager from peer list, and also consolidated the use of <driver> in BlockManager.
7598f91 [Tathagata Das] Minor changes.
03de02d [Tathagata Das] Change replication logic to correctly refetch peers from master on failure and on new worker addition.
d081bf6 [Tathagata Das] Fixed bug in get peers and unit tests to test get-peers and replication under executor churn.
9f0ac9f [Tathagata Das] Modified replication tests to fail on replication bug.
af0c1da [Tathagata Das] Added replication unit tests to BlockManagerSuite
### Problem
The section "Using the shell" in Spark Programming Guide (https://spark.apache.org/docs/latest/programming-guide.html#using-the-shell) says that we can run pyspark REPL through IPython.
But a folloing command does not run IPython but a default Python executable.
```
$ IPYTHON=1 ./bin/pyspark
Python 2.7.8 (default, Jul 2 2014, 10:14:46)
...
```
the spark/bin/pyspark script on the commit b235e01363 decides which executable and options it use folloing way.
1. if PYSPARK_PYTHON unset
* → defaulting to "python"
2. if IPYTHON_OPTS set
* → set IPYTHON "1"
3. some python scripts passed to ./bin/pyspak → run it with ./bin/spark-submit
* out of this issues scope
4. if IPYTHON set as "1"
* → execute $PYSPARK_PYTHON (default: ipython) with arguments $IPYTHON_OPTS
* otherwise execute $PYSPARK_PYTHON
Therefore, when PYSPARK_PYTHON is unset, python is executed though IPYTHON is "1".
In other word, when PYSPARK_PYTHON is unset, IPYTHON_OPS and IPYTHON has no effect on decide which command to use.
PYSPARK_PYTHON | IPYTHON_OPTS | IPYTHON | resulting command | expected command
---- | ---- | ----- | ----- | -----
(unset → defaults to python) | (unset) | (unset) | python | (same)
(unset → defaults to python) | (unset) | 1 | python | ipython
(unset → defaults to python) | an_option | (unset → set to 1) | python an_option | ipython an_option
(unset → defaults to python) | an_option | 1 | python an_option | ipython an_option
ipython | (unset) | (unset) | ipython | (same)
ipython | (unset) | 1 | ipython | (same)
ipython | an_option | (unset → set to 1) | ipython an_option | (same)
ipython | an_option | 1 | ipython an_option | (same)
### Suggestion
The pyspark script should determine firstly whether a user wants to run IPython or other executables.
1. if IPYTHON_OPTS set
* set IPYTHON "1"
2. if IPYTHON has a value "1"
* PYSPARK_PYTHON defaults to "ipython" if not set
3. PYSPARK_PYTHON defaults to "python" if not set
See the pull request for more detailed modification.
Author: cocoatomo <cocoatomo77@gmail.com>
Closes#2554 from cocoatomo/issues/cannot-run-ipython-without-options and squashes the following commits:
d2a9b06 [cocoatomo] [SPARK-3706][PySpark] Use PYTHONUNBUFFERED environment variable instead of -u option
264114c [cocoatomo] [SPARK-3706][PySpark] Remove the sentence about deprecated environment variables
42e02d5 [cocoatomo] [SPARK-3706][PySpark] Replace environment variables used to customize execution of PySpark REPL
10d56fb [cocoatomo] [SPARK-3706][PySpark] Cannot run IPython REPL with IPYTHON set to "1" and PYSPARK_PYTHON unset
This change reorders the replicas returned by
HadoopRDD#getPreferredLocations so that replicas cached by HDFS are at
the start of the list. This requires Hadoop 2.5 or higher; previous
versions of Hadoop do not expose the information needed to determine
whether a replica is cached.
Author: Colin Patrick Mccabe <cmccabe@cloudera.com>
Closes#1486 from cmccabe/SPARK-1767 and squashes the following commits:
338d4f8 [Colin Patrick Mccabe] SPARK-1767: Prefer HDFS-cached replicas when scheduling data-local tasks
FutureAction is the only type exposed through the async APIs, so
for job IDs to be useful they need to be exposed there. The complication
is that some async jobs run more than one job (e.g. takeAsync),
so the exposed ID has to actually be a list of IDs that can actually
change over time. So the interface doesn't look very nice, but...
Change is actually small, I just added a basic test to make sure
it works.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#2337 from vanzin/SPARK-3446 and squashes the following commits:
e166a68 [Marcelo Vanzin] Fix comment.
1fed2bc [Marcelo Vanzin] [SPARK-3446] Expose underlying job ids in FutureAction.
https://issues.apache.org/jira/browse/SPARK-3658
And keep the `CLASS_NOT_FOUND_EXIT_STATUS` and exit message in `SparkSubmit.scala`.
Author: WangTaoTheTonic <barneystinson@aliyun.com>
Author: WangTao <barneystinson@aliyun.com>
Closes#2509 from WangTaoTheTonic/thriftserver and squashes the following commits:
5dcaab2 [WangTaoTheTonic] issue about coupling
8ad9f95 [WangTaoTheTonic] generalization
598e21e [WangTao] take thrift server as a daemon
Non-root user use port 1- 1024 to start jetty server will get the exception " java.net.SocketException: Permission denied", so not use these ports
Author: scwf <wangfei1@huawei.com>
Closes#2610 from scwf/1-1024 and squashes the following commits:
cb8cc76 [scwf] do not use port 1 - 1024
1. broadcast is triggle unexpected
2. fd is leaked in JVM (also leak in parallelize())
3. broadcast is not unpersisted in JVM after RDD is not be used any more.
cc JoshRosen , sorry for these stupid bugs.
Author: Davies Liu <davies.liu@gmail.com>
Closes#2603 from davies/fix_broadcast and squashes the following commits:
080a743 [Davies Liu] fix bugs in broadcast large closure of RDD
Author: Reynold Xin <rxin@apache.org>
Closes#2599 from rxin/SPARK-3747 and squashes the following commits:
a74c04d [Reynold Xin] Added a line of comment explaining NonFatal
0e8d44c [Reynold Xin] [SPARK-3747] TaskResultGetter could incorrectly abort a stage if it cannot get result for a specific task
Author: Reynold Xin <rxin@apache.org>
Closes#2602 from rxin/warning and squashes the following commits:
130186b [Reynold Xin] Remove compiler warning from TaskContext change.
As suggested by mateiz , and because it came up on the mailing list again last week, this attempts to document that ordering of elements is not guaranteed across RDD evaluations in groupBy, zip, and partition-wise RDD methods. Suggestions welcome about the wording, or other methods that need a note.
Author: Sean Owen <sowen@cloudera.com>
Closes#2508 from srowen/SPARK-3356 and squashes the following commits:
b7c96fd [Sean Owen] Undo change to programming guide
ad4aeec [Sean Owen] Don't mention ordering in partition-wise methods, reword description of ordering for zip methods per review, and add similar note to programming guide, which mentions groupByKey (but not zip methods)
fce943b [Sean Owen] Note that ordering of elements is not guaranteed across RDD evaluations in groupBy, zip, and partition-wise RDD methods
When using spark-submit in `cluster` mode to submit a job to a Spark Standalone
cluster, if the JAVA_HOME environment variable was set on the submitting
machine then DriverRunner would attempt to use the submitter's JAVA_HOME to
launch the driver process (instead of the worker's JAVA_HOME), causing the
driver to fail unless the submitter and worker had the same Java location.
This commit fixes this by reading JAVA_HOME from sys.env instead of
command.environment.
Author: Josh Rosen <joshrosen@apache.org>
Closes#2586 from JoshRosen/SPARK-3734 and squashes the following commits:
e9513d9 [Josh Rosen] [SPARK-3734] DriverRunner should not read SPARK_HOME from submitter's environment.
The problem was that the 2nd argument in RemoveBroadcast is not tellMaster! It is "removeFromDriver". Basically when removeFromDriver is not true, we don't report broadcast block removal back to the driver, and then other executors mistakenly think that the executor would still have the block, and try to fetch from it.
cc @tdas
Author: Reynold Xin <rxin@apache.org>
Closes#2588 from rxin/debug and squashes the following commits:
6dab2e3 [Reynold Xin] Don't log random messages.
f430686 [Reynold Xin] Always report broadcast removal back to master.
2a13f70 [Reynold Xin] iii
This changes the way we send MapStatus from executors back to driver for large stages (>2000 tasks). For large stages, we no longer send one byte per block. Instead, we just send the average block size.
This makes large jobs (tens of thousands of tasks) much more reliable since the driver no longer sends huge amount of data.
Author: Reynold Xin <rxin@apache.org>
Closes#2470 from rxin/mapstatus and squashes the following commits:
822ff54 [Reynold Xin] Code review feedback.
3b86f56 [Reynold Xin] Added MimaExclude.
f89d182 [Reynold Xin] Fixed a bug in MapStatus
6a0401c [Reynold Xin] [SPARK-3613] Record only average block size in MapStatus for large stages.
Author: Reynold Xin <rxin@apache.org>
Closes#2581 from rxin/minor-cleanup and squashes the following commits:
736a91b [Reynold Xin] Minor cleanup of code.
Previous key comparison in `ExternalSorter` will get wrong sorting result or exception when key comparison overflows, details can be seen in [SPARK-3032](https://issues.apache.org/jira/browse/SPARK-3032). Here fix this and add a unit test to prove it.
Author: jerryshao <saisai.shao@intel.com>
Closes#2514 from jerryshao/SPARK-3032 and squashes the following commits:
6f3c302 [jerryshao] Improve the unit test according to comments
01911e6 [jerryshao] Change the test to show the contract violate exception
83acb38 [jerryshao] Minor changes according to comments
fa2a08f [jerryshao] Fix key comparison integer overflow introduced sorting exception
Author: Zhang, Liye <liye.zhang@intel.com>
Closes#2572 from liyezhang556520/DAGLogErr and squashes the following commits:
5be2491 [Zhang, Liye] Bugfix: LogErr format in DAGScheduler.scala
Author: Reynold Xin <rxin@apache.org>
Closes#2560 from rxin/TaskContext and squashes the following commits:
9eff95a [Reynold Xin] [SPARK-3543] remaining cleanup work.
Moved `#maybeSpill` in ExternalSorter and EAOM into `Spillable`.
Author: Jim Lim <jim@quixey.com>
Closes#2416 from jimjh/SPARK-2761 and squashes the following commits:
cf8be9a [Jim Lim] SPARK-2761 fix documentation, reorder code
f94d522 [Jim Lim] SPARK-2761 refactor Spillable to simplify sig
e75a24e [Jim Lim] SPARK-2761 use protected over protected[this]
7270e0d [Jim Lim] SPARK-2761 refactor #maybeSpill into Spillable
from [SPARK-3651]
In CoarseGrainedSchedulerBackend, we have:
private val executorActor = new HashMap[String, ActorRef]
private val executorAddress = new HashMap[String, Address]
private val executorHost = new HashMap[String, String]
private val freeCores = new HashMap[String, Int]
private val totalCores = new HashMap[String, Int]
We only ever put / remove stuff from these maps together. It would simplify the code if we consolidate these all into one map as we have done in JobProgressListener in https://issues.apache.org/jira/browse/SPARK-2299.
Author: Dale <tigerquoll@outlook.com>
Closes#2533 from tigerquoll/SPARK-3651 and squashes the following commits:
d1be0a9 [Dale] [SPARK-3651] implemented suggested changes. Changed a reference from executorInfo to executorData to be consistent with other usages
6890663 [Dale] [SPARK-3651] implemented suggested changes
7d671cf [Dale] [SPARK-3651] Grouped variables under a ExecutorDataObject, and reference them via a map entry as they are all retrieved under the same key
This addresses some minor issues in https://github.com/apache/spark/pull/2425
Author: Reynold Xin <rxin@apache.org>
Closes#2557 from rxin/TaskContext and squashes the following commits:
a51e5f6 [Reynold Xin] [SPARK-3543] Clean up Java TaskContext implementation.
Author: Reynold Xin <rxin@apache.org>
Closes#2555 from rxin/cleanup and squashes the following commits:
6add199 [Reynold Xin] Minor cleanup to tighten visibility and remove compilation warning.
...rtByKey() until evaluation.
Author: Erik Erlandson <eerlands@redhat.com>
Closes#1689 from erikerlandson/spark-1021-pr and squashes the following commits:
50b6da6 [Erik Erlandson] use standard getIteratorSize in countAsync
4e334a9 [Erik Erlandson] exception mystery fixed by fixing bug in ComplexFutureAction
b88b5d4 [Erik Erlandson] tweak async actions to use ComplexFutureAction[T] so they handle RangePartitioner sampling job properly
b2b20e8 [Erik Erlandson] Fix bug in exception passing with ComplexFutureAction[T]
ca8913e [Erik Erlandson] RangePartition sampling job -> FutureAction
7143f97 [Erik Erlandson] [SPARK-1021] modify range bounds variable to be thread safe
ac67195 [Erik Erlandson] [SPARK-1021] Defer the data-driven computation of partition bounds in sortByKey() until evaluation.
Author: Prashant Sharma <prashant.s@imaginea.com>
Author: Shashank Sharma <shashank21j@gmail.com>
Closes#2425 from ScrapCodes/SPARK-3543/withTaskContext and squashes the following commits:
8ae414c [Shashank Sharma] CR
ee8bd00 [Prashant Sharma] Added internal API in docs comments.
ddb8cbe [Prashant Sharma] Moved setting the thread local to where TaskContext is instantiated.
a7d5e23 [Prashant Sharma] Added doc comments.
edf945e [Prashant Sharma] Code review git add -A
f716fd1 [Prashant Sharma] introduced thread local for getting the task context.
333c7d6 [Prashant Sharma] Translated Task context from scala to java.
should output detailed host and port in error message
Author: Daoyuan Wang <daoyuan.wang@intel.com>
Closes#2539 from adrian-wang/fetchfail and squashes the following commits:
6c1b1e0 [Daoyuan Wang] shuffle fetch fail output
MapOutputTrackerWorker.mapStatuses is used concurrently, it should be thread-safe. This bug has already been fixed in #1328. Nevertheless, considering #1328 won't be merged soon, I send this trivial fix and hope this issue can be solved soon.
Author: zsxwing <zsxwing@gmail.com>
Closes#1541 from zsxwing/SPARK-2634 and squashes the following commits:
d450053 [zsxwing] SPARK-2634: Change MapOutputTrackerWorker.mapStatuses to ConcurrentHashMap
Author: epahomov <pahomov.egor@gmail.com>
Closes#2537 from epahomov/SPARK-3690 and squashes the following commits:
a0b7de4 [epahomov] [SPARK-3690] Closing shuffle writers we swallow more important exception
leftOuterJoin and rightOuterJoin are already implemented. This patch adds fullOuterJoin.
Author: Aaron Staple <aaron.staple@gmail.com>
Closes#1395 from staple/SPARK-546 and squashes the following commits:
1f5595c [Aaron Staple] Fix python style
7ac0aa9 [Aaron Staple] [SPARK-546] Add full outer join to RDD and DStream.
3b5d137 [Aaron Staple] In JavaPairDStream, make class tag specification in rightOuterJoin consistent with other functions.
31f2956 [Aaron Staple] Fix left outer join documentation comments.
This is a refactored version of the original PR https://github.com/apache/spark/pull/1723 my mubarak
Please take a look andrewor14, mubarak
Author: Mubarak Seyed <mubarak.seyed@gmail.com>
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#2464 from tdas/streaming-callsite and squashes the following commits:
dc54c71 [Tathagata Das] Made changes based on PR comments.
390b45d [Tathagata Das] Fixed minor bugs.
904cd92 [Tathagata Das] Merge remote-tracking branch 'apache-github/master' into streaming-callsite
7baa427 [Tathagata Das] Refactored getCallSite and setCallSite to make it simpler. Also added unit test for DStream creation site.
b9ed945 [Mubarak Seyed] Adding streaming utils
c461cf4 [Mubarak Seyed] Merge remote-tracking branch 'upstream/master'
ceb43da [Mubarak Seyed] Changing default regex function name
8c5d443 [Mubarak Seyed] Merge remote-tracking branch 'upstream/master'
196121b [Mubarak Seyed] Merge remote-tracking branch 'upstream/master'
491a1eb [Mubarak Seyed] Removing streaming visibility from getRDDCreationCallSite in DStream
33a7295 [Mubarak Seyed] Fixing review comments: Merging both setCallSite methods
c26d933 [Mubarak Seyed] Merge remote-tracking branch 'upstream/master'
f51fd9f [Mubarak Seyed] Fixing scalastyle, Regex for Utils.getCallSite, and changing method names in DStream
5051c58 [Mubarak Seyed] Getting return value of compute() into variable and call setCallSite(prevCallSite) only once. Adding return for other code paths (for None)
a207eb7 [Mubarak Seyed] Fixing code review comments
ccde038 [Mubarak Seyed] Removing Utils import from MappedDStream
2a09ad6 [Mubarak Seyed] Changes in Utils.scala for SPARK-1853
1d90cc3 [Mubarak Seyed] Changes for SPARK-1853
5f3105a [Mubarak Seyed] Merge remote-tracking branch 'upstream/master'
70f494f [Mubarak Seyed] Changes for SPARK-1853
1500deb [Mubarak Seyed] Changes in Spark Streaming UI
9d38d3c [Mubarak Seyed] [SPARK-1853] Show Streaming application code context (file, line number) in Spark Stages UI
d466d75 [Mubarak Seyed] Changes for spark streaming UI
`SPARK_DRIVER_MEMORY` was only used to start the `SparkSubmit` JVM, which becomes the driver only in client mode but not cluster mode. In cluster mode, this property is simply not propagated to the worker nodes.
`SPARK_EXECUTOR_MEMORY` is picked up from `SparkContext`, but in cluster mode the driver runs on one of the worker machines, where this environment variable may not be set.
Author: Andrew Or <andrewor14@gmail.com>
Closes#2500 from andrewor14/memory-env-vars and squashes the following commits:
6217b38 [Andrew Or] Respect SPARK_*_MEMORY for cluster mode
...the driver
Author: Sandy Ryza <sandy@cloudera.com>
Closes#2487 from sryza/sandy-spark-3612 and squashes the following commits:
2b7353d [Sandy Ryza] SPARK-3612. Executor shouldn't quit if heartbeat message fails to reach the driver
Addresses the issue in https://issues.apache.org/jira/browse/SPARK-3595, namely saveAsHadoopFile hardcoding the OutputCommitter. This is not ideal when running Spark jobs that write to S3, especially when running them from an EMR cluster where the default OutputCommitter is a DirectOutputCommitter.
Author: Ian Hummel <ian@themodernlife.net>
Closes#2450 from themodernlife/spark-3595 and squashes the following commits:
f37a0e5 [Ian Hummel] Update based on comments from pwendell
a11d9f3 [Ian Hummel] Fix formatting
4359664 [Ian Hummel] Add an example showing usage
8b6be94 [Ian Hummel] Add ability to specify OutputCommitter, espcially useful when writing to an S3 bucket from an EMR cluster
https://issues.apache.org/jira/browse/SPARK-3599
Author: WangTao <barneystinson@aliyun.com>
Author: WangTaoTheTonic <barneystinson@aliyun.com>
Closes#2454 from WangTaoTheTonic/avoidLoadingFrequently and squashes the following commits:
3681182 [WangTao] do not use clone
7dca036 [WangTao] use lazy val instead
2a79f26 [WangTaoTheTonic] Avoid loaing properties file frequently
The included test waits 100 ms after job completion for task completion events to come in so it can verify they have reasonable finish times. Does anyone know a better way to wait on listener events that are expected to come in?
Author: Sandy Ryza <sandy@cloudera.com>
Closes#2440 from sryza/sandy-spark-3574 and squashes the following commits:
c81439b [Sandy Ryza] Fix test failure
b340956 [Sandy Ryza] SPARK-3574. Remove shuffleFinishTime metric
Currently, we serialize the data between JVM and Python case by case manually, this cannot scale to support so many APIs in MLlib.
This patch will try to address this problem by serialize the data using pickle protocol, using Pyrolite library to serialize/deserialize in JVM. Pickle protocol can be easily extended to support customized class.
All the modules are refactored to use this protocol.
Known issues: There will be some performance regression (both CPU and memory, the serialized data increased)
Author: Davies Liu <davies.liu@gmail.com>
Closes#2378 from davies/pickle_mllib and squashes the following commits:
dffbba2 [Davies Liu] Merge branch 'master' of github.com:apache/spark into pickle_mllib
810f97f [Davies Liu] fix equal of matrix
032cd62 [Davies Liu] add more type check and conversion for user_product
bd738ab [Davies Liu] address comments
e431377 [Davies Liu] fix cache of rdd, refactor
19d0967 [Davies Liu] refactor Picklers
2511e76 [Davies Liu] cleanup
1fccf1a [Davies Liu] address comments
a2cc855 [Davies Liu] fix tests
9ceff73 [Davies Liu] test size of serialized Rating
44e0551 [Davies Liu] fix cache
a379a81 [Davies Liu] fix pickle array in python2.7
df625c7 [Davies Liu] Merge commit '154d141' into pickle_mllib
154d141 [Davies Liu] fix autobatchedpickler
44736d7 [Davies Liu] speed up pickling array in Python 2.7
e1d1bfc [Davies Liu] refactor
708dc02 [Davies Liu] fix tests
9dcfb63 [Davies Liu] fix style
88034f0 [Davies Liu] rafactor, address comments
46a501e [Davies Liu] choose batch size automatically
df19464 [Davies Liu] memorize the module and class name during pickleing
f3506c5 [Davies Liu] Merge branch 'master' into pickle_mllib
722dd96 [Davies Liu] cleanup _common.py
0ee1525 [Davies Liu] remove outdated tests
b02e34f [Davies Liu] remove _common.py
84c721d [Davies Liu] Merge branch 'master' into pickle_mllib
4d7963e [Davies Liu] remove muanlly serialization
6d26b03 [Davies Liu] fix tests
c383544 [Davies Liu] classification
f2a0856 [Davies Liu] mllib/regression
d9f691f [Davies Liu] mllib/util
cccb8b1 [Davies Liu] mllib/tree
8fe166a [Davies Liu] Merge branch 'pickle' into pickle_mllib
aa2287e [Davies Liu] random
f1544c4 [Davies Liu] refactor clustering
52d1350 [Davies Liu] use new protocol in mllib/stat
b30ef35 [Davies Liu] use pickle to serialize data for mllib/recommendation
f44f771 [Davies Liu] enable tests about array
3908f5c [Davies Liu] Merge branch 'master' into pickle
c77c87b [Davies Liu] cleanup debugging code
60e4e2f [Davies Liu] support unpickle array.array for Python 2.6
This was introduced in #2449
Author: Andrew Or <andrewor14@gmail.com>
Closes#2452 from andrewor14/standalone-hot-fix and squashes the following commits:
d5190ca [Andrew Or] Put that line in the right place
Author: Victsm <victor.nju@gmail.com>
Author: Min Shen <mshen@linkedin.com>
Closes#2449 from Victsm/SPARK-3560 and squashes the following commits:
918405a [Victsm] Removed the additional space
4502a2a [Min Shen] [SPARK-3560] Fixed setting spark.jars system property in yarn-cluster mode.
(cherry picked from commit 832dff64dd)
Signed-off-by: Andrew Or <andrewor14@gmail.com>
https://issues.apache.org/jira/browse/SPARK-3589
"export CLASSPATH" in spark-class is redundant since same variable is exported before.
We could reuse defined value "isYarnCluster" in SparkSubmit.scala.
Author: WangTaoTheTonic <barneystinson@aliyun.com>
Closes#2445 from WangTaoTheTonic/removeRedundant and squashes the following commits:
6fb6872 [WangTaoTheTonic] remove redundant code
...on
As improvement of https://github.com/apache/spark/pull/1944, we should use more special exit code to represent ClassNotFoundException.
Author: WangTaoTheTonic <barneystinson@aliyun.com>
Closes#2421 from WangTaoTheTonic/classnotfoundExitCode and squashes the following commits:
645a22a [WangTaoTheTonic] Serveral typos to trigger Jenkins
d6ae559 [WangTaoTheTonic] use 101 instead
a2d6465 [WangTaoTheTonic] use 127 instead
fbb232f [WangTaoTheTonic] Using a special exit code instead of 1 to represent ClassNotFoundException
https://issues.apache.org/jira/browse/SPARK-3565
"spark.ports.maxRetries" should be "spark.port.maxRetries". Make the configuration keys in document and code consistent.
Author: WangTaoTheTonic <barneystinson@aliyun.com>
Closes#2427 from WangTaoTheTonic/fixPortRetries and squashes the following commits:
c178813 [WangTaoTheTonic] Use blank lines trigger Jenkins
646f3fe [WangTaoTheTonic] also in SparkBuild.scala
3700dba [WangTaoTheTonic] Fix configuration item not consistent with document
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2428 from sarutak/appid-volatile-modification and squashes the following commits:
c7d890d [Kousuke Saruta] Added volatile modifier to appId field in SparkDeploySchedulerBackend
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2424 from sarutak/display-appid-on-webui and squashes the following commits:
417fe90 [Kousuke Saruta] Added "App ID column" to HistoryPage
I think, this issue is caused by #1106
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2436 from sarutak/SPARK-3571 and squashes the following commits:
7a4deea [Kousuke Saruta] Modified Master.scala to use numWorkersVisited and numWorkersAlive instead of stopPos
4e51e35 [Kousuke Saruta] Modified Master to prevent from 0 divide
4817ecd [Kousuke Saruta] Brushed up previous change
71e84b6 [Kousuke Saruta] Modified Master to enable schedule normally
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2408 from sarutak/resolve-resource-leak-issue and squashes the following commits:
074781d [Kousuke Saruta] Modified SuffleBlockFetcherIterator
5f63f67 [Kousuke Saruta] Move metrics increment logic and debug logging outside try block
b37231a [Kousuke Saruta] Modified FileSegmentManagedBuffer#nioByteBuffer to check null or not before invoking channel.close
bf29d4a [Kousuke Saruta] Modified FileSegment to close channel
SchemaRDD overrides RDD functions, including collect, count, and take, with optimized versions making use of the query optimizer. The java and python interface classes wrapping SchemaRDD need to ensure the optimized versions are called as well. This patch overrides relevant calls in the python and java interfaces with optimized versions.
Adds a new Row serialization pathway between python and java, based on JList[Array[Byte]] versus the existing RDD[Array[Byte]]. I wasn’t overjoyed about doing this, but I noticed that some QueryPlans implement optimizations in executeCollect(), which outputs an Array[Row] rather than the typical RDD[Row] that can be shipped to python using the existing serialization code. To me it made sense to ship the Array[Row] over to python directly instead of converting it back to an RDD[Row] just for the purpose of sending the Rows to python using the existing serialization code.
Author: Aaron Staple <aaron.staple@gmail.com>
Closes#1592 from staple/SPARK-2314 and squashes the following commits:
89ff550 [Aaron Staple] Merge with master.
6bb7b6c [Aaron Staple] Fix typo.
b56d0ac [Aaron Staple] [SPARK-2314][SQL] Override count in JavaSchemaRDD, forwarding to SchemaRDD's count.
0fc9d40 [Aaron Staple] Fix comment typos.
f03cdfa [Aaron Staple] [SPARK-2314][SQL] Override collect and take in sql.py, forwarding to SchemaRDD's collect.
Short version: NetworkInterface.getNetworkInterfaces returns ifs in reverse order compared to ifconfig output. It may pick up ip address associated with tun0 or virtual network interface.
See [SPARK_3040](https://issues.apache.org/jira/browse/SPARK-3040) for more detail
Author: Ye Xianjin <advancedxy@gmail.com>
Closes#1946 from advancedxy/SPARK-3040 and squashes the following commits:
f33f6b2 [Ye Xianjin] add windows support
087a785 [Ye Xianjin] reverse the Networkinterface.getNetworkInterfaces output order to get a more proper local ip address.
Pyrolite can not unpickle array.array which pickled by Python 2.6, this patch fix it by extend Pyrolite.
There is a bug in Pyrolite when unpickle array of float/double, this patch workaround it by reverse the endianness for float/double. This workaround should be removed after Pyrolite have a new release to fix this issue.
I had send an PR to Pyrolite to fix it: https://github.com/irmen/Pyrolite/pull/11
Author: Davies Liu <davies.liu@gmail.com>
Closes#2365 from davies/pickle and squashes the following commits:
f44f771 [Davies Liu] enable tests about array
3908f5c [Davies Liu] Merge branch 'master' into pickle
c77c87b [Davies Liu] cleanup debugging code
60e4e2f [Davies Liu] support unpickle array.array for Python 2.6
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2380 from sarutak/SPARK-3518 and squashes the following commits:
8a1464e [Kousuke Saruta] Replaced a variable with simple field reference
c660fbc [Kousuke Saruta] Removed useless statement in JsonProtocol.scala
Aggregate the number of bytes spilled into disks during aggregation or sorting, show them in Web UI.
This patch is blocked by SPARK-3465. (It includes a fix for that).
Author: Davies Liu <davies.liu@gmail.com>
Closes#2336 from davies/metrics and squashes the following commits:
e37df38 [Davies Liu] remove outdated comments
1245eb7 [Davies Liu] remove the temporary fix
ebd2f43 [Davies Liu] Merge branch 'master' into metrics
7e4ad04 [Davies Liu] Merge branch 'master' into metrics
fbe9029 [Davies Liu] show spilled bytes in Python in web ui
Reuse Python worker to avoid the overhead of fork() Python process for each tasks. It also tracks the broadcasts for each worker, avoid sending repeated broadcasts.
This can reduce the time for dummy task from 22ms to 13ms (-40%). It can help to reduce the latency for Spark Streaming.
For a job with broadcast (43M after compress):
```
b = sc.broadcast(set(range(30000000)))
print sc.parallelize(range(24000), 100).filter(lambda x: x in b.value).count()
```
It will finish in 281s without reused worker, and it will finish in 65s with reused worker(4 CPUs). After reusing the worker, it can save about 9 seconds for transfer and deserialize the broadcast for each tasks.
It's enabled by default, could be disabled by `spark.python.worker.reuse = false`.
Author: Davies Liu <davies.liu@gmail.com>
Closes#2259 from davies/reuse-worker and squashes the following commits:
f11f617 [Davies Liu] Merge branch 'master' into reuse-worker
3939f20 [Davies Liu] fix bug in serializer in mllib
cf1c55e [Davies Liu] address comments
3133a60 [Davies Liu] fix accumulator with reused worker
760ab1f [Davies Liu] do not reuse worker if there are any exceptions
7abb224 [Davies Liu] refactor: sychronized with itself
ac3206e [Davies Liu] renaming
8911f44 [Davies Liu] synchronized getWorkerBroadcasts()
6325fc1 [Davies Liu] bugfix: bid >= 0
e0131a2 [Davies Liu] fix name of config
583716e [Davies Liu] only reuse completed and not interrupted worker
ace2917 [Davies Liu] kill python worker after timeout
6123d0f [Davies Liu] track broadcasts for each worker
8d2f08c [Davies Liu] reuse python worker
... that expose a stop() lifecycle method. This doesn't add `AutoCloseable`, which is Java 7+ only. But it should be possible to use try-with-resources on a `Closeable` in Java 7, as long as the `close()` does not throw a checked exception, and these don't. Q.E.D.
Author: Sean Owen <sowen@cloudera.com>
Closes#2346 from srowen/SPARK-3470 and squashes the following commits:
612c21d [Sean Owen] Add Closeable / close() to Java context objects that expose a stop() lifecycle method
This is necessary because we rely on this callback interface to clean resources up. The old behavior would lead to resource leaks.
Note that this also changes the fault semantics of TaskCompletionListener. Previously failures in TaskCompletionListeners would result in the task being reported immediately. With this change, we report the exception at the end, and the reported exception is a TaskCompletionListenerException that contains all the exception messages.
Author: Reynold Xin <rxin@apache.org>
Closes#2343 from rxin/taskcontext-callback and squashes the following commits:
a3845b2 [Reynold Xin] Mark TaskCompletionListenerException as private[spark].
ac5baea [Reynold Xin] Removed obsolete comment.
aa68ea4 [Reynold Xin] Throw an exception if task completion callback fails.
29b6162 [Reynold Xin] oops compilation failed.
1cb444d [Reynold Xin] [SPARK-3469] Call all TaskCompletionListeners even if some fail.
...s
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1934 from sryza/sandy-spark-3014 and squashes the following commits:
ae19cc1 [Sandy Ryza] SPARK-3014. Log a more informative messages in a couple failure scenarios
Before overwrite t.taskMetrics, take a deepcopy of it.
Author: Davies Liu <davies.liu@gmail.com>
Closes#2338 from davies/fix_metric and squashes the following commits:
a5cdb63 [Davies Liu] Merge branch 'master' into fix_metric
7c879e0 [Davies Liu] add more comments
754b5b8 [Davies Liu] copy taskMetrics only when isLocal is true
5ca26dc [Davies Liu] fix task metrics aggregation in local mode
Changes logging from
```
14/09/05 02:01:08 INFO SecurityManager: Changing view acls to: aash,
14/09/05 02:01:08 INFO SecurityManager: Changing modify acls to: aash,
14/09/05 02:01:08 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(aash, ); users with modify permissions: Set(aash, )
```
to
```
14/09/05 02:28:28 INFO SecurityManager: Changing view acls to: aash
14/09/05 02:28:28 INFO SecurityManager: Changing modify acls to: aash
14/09/05 02:28:28 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(aash); users with modify permissions: Set(aash)
```
Note that the first set of logs have a Set of size 2 containing "aash" and the empty string ""
cc tgravescs
Author: Andrew Ash <andrew@andrewash.com>
Closes#2286 from ash211/empty-default-acl and squashes the following commits:
18cc612 [Andrew Ash] Use .isEmpty instead of ==""
cf973a1 [Andrew Ash] Don't include the empty string "" as a defaultAclUser
We currently open many ephemeral ports during the tests, and as a result we occasionally can't bind to new ones. This has caused the `DriverSuite` and the `SparkSubmitSuite` to fail intermittently.
By disabling the `SparkUI` when it's not needed, we already cut down on the number of ports opened significantly, on the order of the number of `SparkContexts` ever created. We must keep it enabled for a few tests for the UI itself, however.
Author: Andrew Or <andrewor14@gmail.com>
Closes#2363 from andrewor14/disable-ui-for-tests and squashes the following commits:
332a7d5 [Andrew Or] No need to set spark.ui.port to 0 anymore
30c93a2 [Andrew Or] Simplify streaming UISuite
a431b84 [Andrew Or] Fix streaming test failures
8f5ae53 [Andrew Or] Fix no new line at the end
29c9b5b [Andrew Or] Disable SparkUI for tests
If the waiting driver array is too big, the drivers in it will be dispatched to the first worker we get(if it has enough resources), with or without the Randomization.
We should do randomization every time we dispatch a driver, in order to better balance drivers.
Author: WangTaoTheTonic <barneystinson@aliyun.com>
Author: WangTao <barneystinson@aliyun.com>
Closes#1106 from WangTaoTheTonic/fixBalanceDrivers and squashes the following commits:
d1a928b [WangTaoTheTonic] Minor adjustment
b6560cf [WangTaoTheTonic] solve the shuffle problem for HashSet
f674e59 [WangTaoTheTonic] add comment and minor fix
2835929 [WangTao] solve the failed test and avoid filtering
2ca3091 [WangTao] fix checkstyle
bc91bb1 [WangTao] Avoid shuffle every time we schedule the driver using round robin
bbc7087 [WangTaoTheTonic] Optimize the schedule in Master
Author: Prashant Sharma <prashant.s@imaginea.com>
Closes#2331 from ScrapCodes/compilation-warn and squashes the following commits:
44c1e76 [Prashant Sharma] Minor - Fix trivial compilation warnings.
https://issues.apache.org/jira/browse/SPARK-3193
I noticed that sometimes pr tests failed due to the Process exitcode != 0,refer to
https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/18688/consoleFullhttps://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/19118/consoleFull
[info] SparkSubmitSuite:
[info] - prints usage on empty input
[info] - prints usage with only --help
[info] - prints error with unrecognized options
[info] - handle binary specified but not class
[info] - handles arguments with --key=val
[info] - handles arguments to user program
[info] - handles arguments to user program with name collision
[info] - handles YARN cluster mode
[info] - handles YARN client mode
[info] - handles standalone cluster mode
[info] - handles standalone client mode
[info] - handles mesos client mode
[info] - handles confs with flag equivalents
[info] - launch simple application with spark-submit *** FAILED ***
[info] org.apache.spark.SparkException: Process List(./bin/spark-submit, --class, org.apache.spark.deploy.SimpleApplicationTest, --name, testApp, --master, local, file:/tmp/1408854098404-0/testJar-1408854098404.jar) exited with code 1
[info] at org.apache.spark.util.Utils$.executeAndGetOutput(Utils.scala:872)
[info] at org.apache.spark.deploy.SparkSubmitSuite.runSparkSubmit(SparkSubmitSuite.scala:311)
[info] at org.apache.spark.deploy.SparkSubmitSuite$$anonfun$14.apply$mcV$sp(SparkSubmitSuite.scala:291)
[info] at org.apache.spark.deploy.SparkSubmitSuite$$anonfun$14.apply(SparkSubmitSuite.scala:284)
[info] at org.apacSpark assembly has been built with Hive, including Datanucleus jars on classpath
this PR output the process error info when failed, it can be helpful for diagnosis.
Author: scwf <wangfei1@huawei.com>
Closes#2108 from scwf/output-test-error-info and squashes the following commits:
0c48082 [scwf] minor fix according to comments
563fde1 [scwf] output errer info when Process exitcode not zero
Introduces a LOADING -> RUNNING ApplicationState transition and prevents Master from removing an Application with RUNNING Executors.
Two basic changes: 1) Instead of allowing MAX_NUM_RETRY abnormal Executor exits over the entire lifetime of the Application, allow that many since any Executor successfully began running the Application; 2) Don't remove the Application while Master still thinks that there are RUNNING Executors.
This should be fine as long as the ApplicationInfo doesn't believe any Executors are forever RUNNING when they are not. I think that any non-RUNNING Executors will eventually no longer be RUNNING in Master's accounting, but another set of eyes should confirm that. This PR also doesn't try to detect which nodes have gone rogue or to kill off bad Workers, so repeatedly failing Executors will continue to fail and fill up log files with failure reports as long as the Application keeps running.
Author: Mark Hamstra <markhamstra@gmail.com>
Closes#1360 from markhamstra/SPARK-2425 and squashes the following commits:
f099c0b [Mark Hamstra] Reuse appInfo
b2b7b25 [Mark Hamstra] Moved 'Application failed' logging
bdd0928 [Mark Hamstra] switched to string interpolation
1dd591b [Mark Hamstra] SPARK-2425 introduce LOADING -> RUNNING ApplicationState transition and prevent Master from removing Application with RUNNING Executors
This pull request creates a new BlockTransferService interface for block fetch/upload and refactors the existing ConnectionManager to implement BlockTransferService (NioBlockTransferService).
Most of the changes are simply moving code around. The main class to inspect is ShuffleBlockFetcherIterator.
Review guide:
- Most of the ConnectionManager code is now in network.cm package
- ManagedBuffer is a new buffer abstraction backed by several different implementations (file segment, nio ByteBuffer, Netty ByteBuf)
- BlockTransferService is the main internal interface introduced in this PR
- NioBlockTransferService implements BlockTransferService and replaces the old BlockManagerWorker
- ShuffleBlockFetcherIterator replaces the told BlockFetcherIterator to use the new interface
TODOs that should be separate PRs:
- Implement NettyBlockTransferService
- Finalize the API/semantics for ManagedBuffer.release()
Author: Reynold Xin <rxin@apache.org>
Closes#2240 from rxin/blockTransferService and squashes the following commits:
64cd9d7 [Reynold Xin] Merge branch 'master' into blockTransferService
1dfd3d7 [Reynold Xin] Limit the length of the FileInputStream.
1332156 [Reynold Xin] Fixed style violation from refactoring.
2960c93 [Reynold Xin] Added ShuffleBlockFetcherIteratorSuite.
e29c721 [Reynold Xin] Updated comment for ShuffleBlockFetcherIterator.
8a1046e [Reynold Xin] Code review feedback:
2c6b1e1 [Reynold Xin] Removed println in test cases.
2a907e4 [Reynold Xin] Merge branch 'master' into blockTransferService-merge
07ccf0d [Reynold Xin] Added init check to CMBlockTransferService.
98c668a [Reynold Xin] Added failure handling and fixed unit tests.
ae05fcd [Reynold Xin] Updated tests, although DistributedSuite is hanging.
d8d595c [Reynold Xin] Merge branch 'master' of github.com:apache/spark into blockTransferService
9ef279c [Reynold Xin] Initial refactoring to move ConnectionManager to use the BlockTransferService.
I didn't add this to the transformations list in the docs because it's kind of obscure, but would be happy to do so if others think it would be helpful.
Author: Sandy Ryza <sandy@cloudera.com>
Closes#2274 from sryza/sandy-spark-2978 and squashes the following commits:
4a5332a [Sandy Ryza] Fix Java test
c04b447 [Sandy Ryza] Fix Python doc and add back deleted code
433ad5b [Sandy Ryza] Add Java test
4c25a54 [Sandy Ryza] Add s at the end and a couple other fixes
9b0ba99 [Sandy Ryza] Fix compilation
36e0571 [Sandy Ryza] Fix import ordering
48c12c2 [Sandy Ryza] Add Java version and additional doc
e5381cd [Sandy Ryza] Fix python style warnings
f147634 [Sandy Ryza] SPARK-2978. Transformation with MR shuffle semantics
Sort-based shuffle has lower memory usage and seems to outperform hash-based in almost all of our testing.
Author: Reynold Xin <rxin@apache.org>
Closes#2178 from rxin/sort-shuffle and squashes the following commits:
713d341 [Reynold Xin] Fixed test failures by setting spark.shuffle.compress to the same value as spark.shuffle.spill.compress.
85165e6 [Reynold Xin] Fixed a comment typo.
aa0d372 [Reynold Xin] [SPARK-3280] Made sort-based shuffle the default implementation
This resolves https://issues.apache.org/jira/browse/SPARK-3394
Author: Eric Liang <ekl@google.com>
Closes#2264 from ericl/spark-3394 and squashes the following commits:
c87355b [Eric Liang] refactor
bfb6140 [Eric Liang] change RDD takeOrdered instead
7a51528 [Eric Liang] fix takeordered when limit = 0
Previously parent stages had higher stage id, but parent stages are executed first. This pull request changes the behavior so parent stages would have lower stage id.
For example, command:
```scala
sc.parallelize(1 to 10).map(x=>(x,x)).reduceByKey(_+_).count
```
breaks down into 2 stages.
The old web UI:
Web UI with this patch:
Author: Reynold Xin <rxin@apache.org>
Closes#2273 from rxin/lower-stage-id and squashes the following commits:
abbb4c6 [Reynold Xin] Fixed SparkListenerSuite.
0e02379 [Reynold Xin] Updated DAGSchedulerSuite.
54ccea3 [Reynold Xin] [SPARK-3353] parent stage should have lower stage id.
Author: GuoQiang Li <witgo@qq.com>
Closes#2175 from witgo/SPARK-3273 and squashes the following commits:
cf9c65a [GuoQiang Li] We should read the version information from the same place
2a44e2f [GuoQiang Li] The spark version in the welcome message of pyspark is not correct
Instead of jumping straight from 1 partition to all partitions, do exponential
growth and double the number of partitions to attempt each time instead.
Fix proposed by Paul Nepywoda
Author: Andrew Ash <andrew@andrewash.com>
Closes#2117 from ash211/SPARK-3211 and squashes the following commits:
8b2299a [Andrew Ash] Quadruple instead of double for a minor speedup
e5f7e4d [Andrew Ash] Update comment to better reflect what we're doing
09a27f7 [Andrew Ash] Update PySpark to be less OOM-prone as well
3a156b8 [Andrew Ash] SPARK-3211 .take() is OOM-prone with empty partitions
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2138 from sarutak/SPARK-3233 and squashes the following commits:
c0205b7 [Kousuke Saruta] Merge branch 'SPARK-3233' of github.com:sarutak/spark into SPARK-3233
064679d [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-3233
d3005fd [Kousuke Saruta] Modified Class definition format of BlockManagerMaster
039b747 [Kousuke Saruta] Modified style
889e2d1 [Kousuke Saruta] Modified BlockManagerMaster to be able to be past isDriver flag
4da8535 [Kousuke Saruta] Modified BlockManagerMaster#stop to send StopBlockManagerMaster message when sender is Driver
6518c3a [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-3233
d5ab19a [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-3233
6bce25c [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-3233
6058a58 [Kousuke Saruta] Modified Executor not to invoke SparkEnv#stop in local mode
e5ad9d3 [Kousuke Saruta] Modified Executor to stop SparnEnv at the end of itself
run test with the master branch with this command when mesos native lib is set
sbt/sbt -Phive "test-only org.apache.spark.SparkContextSchedulerCreationSuite"
get this error:
[info] SparkContextSchedulerCreationSuite:
[info] - bad-master
[info] - local
[info] - local-*
[info] - local-n
[info] - local--n-failures
[info] - local-n-failures
[info] - bad-local-n
[info] - bad-local-n-failures
[info] - local-default-parallelism
[info] - simr
[info] - local-cluster
[info] - yarn-cluster
[info] - yarn-standalone
[info] - yarn-client
[info] - mesos fine-grained
[info] - mesos coarse-grained ** FAILED ***
[info] Executor Spark home `spark.mesos.executor.home` is not set!
Since `executorSparkHome` only used in `createCommand`, move `val executorSparkHome...` to `createCommand` to fix this issue.
Author: scwf <wangfei1@huawei.com>
Author: wangfei <wangfei_hello@126.com>
Closes#2199 from scwf/SparkContextSchedulerCreationSuite and squashes the following commits:
ef1de22 [scwf] fix code fomate
19d26f3 [scwf] fix conflict
d9a8a60 [wangfei] fix SparkContextSchedulerCreationSuite test error
In the method `newFileGroup` of class `FileShuffleBlockManager`, the parameters for creating new `ShuffleFileGroup` object is in wrong order.
Because in current codes, the parameters `shuffleId` and `fileId` are not used. So it doesn't cause problem now. However it should be corrected for readability and avoid future problem.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#2235 from viirya/correct_shufflefilegroup_params and squashes the following commits:
fe72567 [Liang-Chi Hsieh] Do correct parameters for ShuffleFileGroup.
This is causing the event logs to include a file called SPARK_VERSION_1.0.0, which is not accurate.
Author: Andrew Or <andrewor14@gmail.com>
Author: andrewor14 <andrewor14@gmail.com>
Closes#2255 from andrewor14/spark-version and squashes the following commits:
1fbdfe9 [andrewor14] Snapshot
805a1c8 [Andrew Or] JK. Update Spark version to 1.2.0 instead.
bffbaab [Andrew Or] Update Spark version to 1.1.0
This change exposes the application ID generated by the Spark Master, Mesos or Yarn
via the SparkListenerApplicationStart event. It then uses that information to expose the
application via its ID in the history server, instead of using the internal directory name
generated by the event logger as an application id. This allows someone who knows
the application ID to easily figure out the URL for the application's entry in the HS, aside
from looking better.
In Yarn mode, this is used to generate a direct link from the RM application list to the
Spark history server entry (thus providing a fix for SPARK-2150).
Note this sort of assumes that the different managers will generate app ids that are
sufficiently different from each other that clashes will not occur.
Author: Marcelo Vanzin <vanzin@cloudera.com>
This patch had conflicts when merged, resolved by
Committer: Andrew Or <andrewor14@gmail.com>
Closes#1218 from vanzin/yarn-hs-link-2 and squashes the following commits:
2d19f3c [Marcelo Vanzin] Review feedback.
6706d3a [Marcelo Vanzin] Implement applicationId() in base classes.
56fe42e [Marcelo Vanzin] Fix cluster mode history address, plus a cleanup.
44112a8 [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
8278316 [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
a86bbcf [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
a0056e6 [Marcelo Vanzin] Unbreak test.
4b10cfd [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
cb0cab2 [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
25f2826 [Marcelo Vanzin] Add MIMA excludes.
f0ba90f [Marcelo Vanzin] Use BufferedIterator.
c90a08d [Marcelo Vanzin] Remove unused code.
3f8ec66 [Marcelo Vanzin] Review feedback.
21aa71b [Marcelo Vanzin] Fix JSON test.
b022bae [Marcelo Vanzin] Undo SparkContext cleanup.
c6d7478 [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
4e3483f [Marcelo Vanzin] Fix test.
57517b8 [Marcelo Vanzin] Review feedback. Mostly, more consistent use of Scala's Option.
311e49d [Marcelo Vanzin] Merge branch 'master' into yarn-hs-link-2
d35d86f [Marcelo Vanzin] Fix yarn backend after rebase.
36dc362 [Marcelo Vanzin] Don't use Iterator::takeWhile().
0afd696 [Marcelo Vanzin] Wait until master responds before returning from start().
abc4697 [Marcelo Vanzin] Make FsHistoryProvider keep a map of applications by id.
26b266e [Marcelo Vanzin] Use Mesos framework ID as Spark application ID.
b3f3664 [Marcelo Vanzin] [yarn] Make the RM link point to the app direcly in the HS.
2fb7de4 [Marcelo Vanzin] Expose the application ID in the ApplicationStart event.
ed10348 [Marcelo Vanzin] Expose application id to spark context.
These are not used by the UI but are useful when analysing the
logs from a spark job.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#654 from vanzin/bm-event-tstamp and squashes the following commits:
d5d6e66 [Marcelo Vanzin] Fix tests.
ec06218 [Marcelo Vanzin] Review feedback.
f134dbc [Marcelo Vanzin] Merge branch 'master' into bm-event-tstamp
b495b7c [Marcelo Vanzin] Merge branch 'master' into bm-event-tstamp
7d2fe9e [Marcelo Vanzin] Review feedback.
d6f381c [Marcelo Vanzin] Update tests added after patch was created.
45e3bf8 [Marcelo Vanzin] Fix unit test after merge.
b37a10f [Marcelo Vanzin] Use === in test assertions.
ef72824 [Marcelo Vanzin] Handle backwards compatibility with 1.0.0.
aca1151 [Marcelo Vanzin] Fix unit test to check new fields.
efdda8e [Marcelo Vanzin] Add timestamps to block manager events.
RDD.countApproxDistinct(relativeSD=0.05):
:: Experimental ::
Return approximate number of distinct elements in the RDD.
The algorithm used is based on streamlib's implementation of
"HyperLogLog in Practice: Algorithmic Engineering of a State
of The Art Cardinality Estimation Algorithm", available
<a href="http://dx.doi.org/10.1145/2452376.2452456">here</a>.
This support all the types of objects, which is supported by
Pyrolite, nearly all builtin types.
param relativeSD Relative accuracy. Smaller values create
counters that require more space.
It must be greater than 0.000017.
>>> n = sc.parallelize(range(1000)).map(str).countApproxDistinct()
>>> 950 < n < 1050
True
>>> n = sc.parallelize([i % 20 for i in range(1000)]).countApproxDistinct()
>>> 18 < n < 22
True
Author: Davies Liu <davies.liu@gmail.com>
Closes#2142 from davies/countApproxDistinct and squashes the following commits:
e20da47 [Davies Liu] remove the correction in Python
c38c4e4 [Davies Liu] fix doc tests
2ab157c [Davies Liu] fix doc tests
9d2565f [Davies Liu] add commments and link for hash collision correction
d306492 [Davies Liu] change range of hash of tuple to [0, maxint]
ded624f [Davies Liu] calculate hash in Python
4cba98f [Davies Liu] add more tests
a85a8c6 [Davies Liu] Merge branch 'master' into countApproxDistinct
e97e342 [Davies Liu] add countApproxDistinct()
...job fails while reading from Hadoop
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1956 from sryza/sandy-spark-3052 and squashes the following commits:
815813a [Sandy Ryza] SPARK-3052. Misleading and spurious FileSystem closed errors whenever a job fails while reading from Hadoop
This PR adds the async actions to the Java API. User can call these async actions to get the FutureAction and use JobWaiter (for SimpleFutureAction) to retrieve job Id.
Author: lirui <rui.li@intel.com>
Closes#2176 from lirui-intel/SPARK-2636 and squashes the following commits:
ccaafb7 [lirui] SPARK-2636: fix java doc
5536d55 [lirui] SPARK-2636: mark the async API as experimental
e2e01d5 [lirui] SPARK-2636: add mima exclude
0ca320d [lirui] SPARK-2636: fix method name & javadoc
3fa39f7 [lirui] SPARK-2636: refine the patch
af4f5d9 [lirui] SPARK-2636: remove unused imports
843276c [lirui] SPARK-2636: only keep foreachAsync in the java API
fbf5744 [lirui] SPARK-2636: add more async actions for java api
1b25abc [lirui] SPARK-2636: expose some fields in JobWaiter
d09f732 [lirui] SPARK-2636: fix build
eb1ee79 [lirui] SPARK-2636: change some parameters in SimpleFutureAction to member field
6e2b87b [lirui] SPARK-2636: add java API for async actions
This also enables supporting broadcast variables larger than 2G.
Author: Reynold Xin <rxin@apache.org>
Closes#2054 from rxin/ByteArrayChunkOutputStream and squashes the following commits:
618d9c8 [Reynold Xin] Code review.
93f5a51 [Reynold Xin] Added comments.
ee88e73 [Reynold Xin] to -> until
bbd1cb1 [Reynold Xin] Renamed a variable.
36f4d01 [Reynold Xin] Sort imports.
8f1a8eb [Reynold Xin] [SPARK-3135] Created ByteArrayChunkOutputStream and used it to avoid memory copy in TorrentBroadcast.
Rather than specifying the path to SparkFiles we need to use the filename.
Author: Holden Karau <holden@pigscanfly.ca>
Closes#2210 from holdenk/SPARK-3318-documentation-for-addfiles-should-say-to-use-file-not-path and squashes the following commits:
a25d27a [Holden Karau] Update the JavaSparkContext addFile method to be clear about using fileName with SparkFiles as well
0ebcb05 [Holden Karau] Documentation update in addFile on how to use SparkFiles.get to specify filename rather than path
Different places in the code were instantiating Configuration / YarnConfiguration objects in different ways. This could lead to confusion for people who actually expected "spark.hadoop.*" options to end up in the configs used by Spark code, since that would only happen for the SparkContext's config.
This change modifies most places to use SparkHadoopUtil to initialize configs, and make that method do the translation that previously was only done inside SparkContext.
The places that were not changed fall in one of the following categories:
- Test code where this doesn't really matter
- Places deep in the code where plumbing SparkConf would be too difficult for very little gain
- Default values for arguments - since the caller can provide their own config in that case
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#1843 from vanzin/SPARK-2889 and squashes the following commits:
52daf35 [Marcelo Vanzin] Merge branch 'master' into SPARK-2889
f179013 [Marcelo Vanzin] Merge branch 'master' into SPARK-2889
51e71cf [Marcelo Vanzin] Add test to ensure that overriding Yarn configs works.
53f9506 [Marcelo Vanzin] Add DeveloperApi annotation.
3d345cb [Marcelo Vanzin] Restore old method for backwards compat.
fc45067 [Marcelo Vanzin] Merge branch 'master' into SPARK-2889
0ac3fdf [Marcelo Vanzin] Merge branch 'master' into SPARK-2889
3f26760 [Marcelo Vanzin] Compilation fix.
f16cadd [Marcelo Vanzin] Initialize config in SparkHadoopUtil.
b8ab173 [Marcelo Vanzin] Update Utils API to take a Configuration argument.
1e7003f [Marcelo Vanzin] Replace explicit Configuration instantiation with SparkHadoopUtil.
By Hiding the shuffleblockmanager behind Shufflemanager, we decouple the shuffle data's block mapping management work from Diskblockmananger. This give a more clear interface and more easy for other shuffle manager to implement their own block management logic. the jira ticket have more details.
Author: Raymond Liu <raymond.liu@intel.com>
Closes#1241 from colorant/shuffle and squashes the following commits:
0e01ae3 [Raymond Liu] Move ShuffleBlockmanager behind shuffleManager
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2200 from sarutak/SPARK-3305 and squashes the following commits:
3cbd6ee [Kousuke Saruta] Removed unused import from classes related to UI
This is a simplified fix for SPARK-1912.
Author: Reynold Xin <rxin@apache.org>
Closes#2179 from rxin/SPARK-1912 and squashes the following commits:
b2f0e9e [Reynold Xin] Fix unit tests.
a8eddfe [Reynold Xin] [SPARK-1912] Lazily initialize buffers for local shuffle blocks.
**Summary of the changes**
The bulk of this PR is comprised of tests and documentation; the actual fix is really just adding 1 line of code (see `BlockObjectWriter.scala`). We currently do not run the `External*` test suites with different compression codecs, and this would have caught the bug reported in [SPARK-3277](https://issues.apache.org/jira/browse/SPARK-3277). This PR extends the existing code to test spilling using all compression codecs known to Spark, including `LZ4`.
**The bug itself**
In `DiskBlockObjectWriter`, we only report the shuffle bytes written before we close the streams. With `LZ4`, all the bytes written reported by our metrics were 0 because `flush()` was not taking effect for some reason. In general, compression codecs may write additional bytes to the file after we call `close()`, and so we must also capture those bytes in our shuffle write metrics.
Thanks mridulm and pwendell for help with debugging.
Author: Andrew Or <andrewor14@gmail.com>
Author: Patrick Wendell <pwendell@gmail.com>
Closes#2187 from andrewor14/fix-lz4-spilling and squashes the following commits:
1b54bdc [Andrew Or] Speed up tests by not compressing everything
1c4624e [Andrew Or] Merge branch 'master' of github.com:apache/spark into fix-lz4-spilling
6b2e7d1 [Andrew Or] Fix compilation error
92e251b [Patrick Wendell] Better documentation for BlockObjectWriter.
a1ad536 [Andrew Or] Fix tests
089593f [Andrew Or] Actually fix SPARK-3277 (tests still fail)
4bbcf68 [Andrew Or] Update tests to actually test all compression codecs
b264a84 [Andrew Or] ExternalAppendOnlyMapSuite code style fixes (minor)
1bfa743 [Andrew Or] Add more information to assert for better debugging
Netty functionality will be added back in subsequent PRs by using the BlockTransferService interface.
Author: Reynold Xin <rxin@apache.org>
Closes#2181 from rxin/SPARK-3281 and squashes the following commits:
5494b0e [Reynold Xin] Fix extra port.
ff6d1e1 [Reynold Xin] [SPARK-3281] Remove Netty specific code in BlockManager.
The executors and the driver may not share the same Spark home. There is currently one way to set the executor side Spark home in Mesos, through setting `spark.home`. However, this is neither documented nor intuitive. This PR adds a more specific config `spark.mesos.executor.home` and exposes this to the user.
liancheng tnachen
Author: Andrew Or <andrewor14@gmail.com>
Closes#2166 from andrewor14/mesos-spark-home and squashes the following commits:
b87965e [Andrew Or] Merge branch 'master' of github.com:apache/spark into mesos-spark-home
f6abb2e [Andrew Or] Document spark.mesos.executor.home
ca7846d [Andrew Or] Add more specific configuration for executor Spark home in Mesos
This is another try after #2145 to fix [SPARK-2608](https://issues.apache.org/jira/browse/SPARK-2608).
The basic idea is to pass `extraJavaOpts` and `extraLibraryPath` together via environment variable `SPARK_EXECUTOR_OPTS`. This variable is recognized by `spark-class` and not used anywhere else. In this way, we still launch Mesos executors with `spark-class`/`spark-executor`, but avoids the executor side Spark home issue.
Quoted string with spaces is not allowed in either `extraJavaOpts` or `extraLibraryPath` when using Spark over Mesos. The reason is that Mesos passes the whole command line as a single string argument to `sh -c` to start the executor, and this makes shell string escaping non-trivial to handle. This should be fixed in a later release.
Classes in package `org.apache.spark.deploy` shouldn't be used as they assume Spark is deployed in standalone mode, and give wrong executor side Spark home directory. Please refer to comments in #2145 for more details.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes#2161 from liancheng/mesos-fix-with-env-var and squashes the following commits:
ba59190 [Cheng Lian] Added fine grained Mesos executor support
1174076 [Cheng Lian] Draft fix for CoarseMesosSchedulerBackend
(cherry picked from commit 935bffe3bf)
Signed-off-by: Reynold Xin <rxin@apache.org>
The issue happens when Spark is run standalone on a cluster.
When master and driver fall simultaneously on one node in a cluster, master tries to recover its state and restart spark driver.
While restarting driver, it falls with NPE exception (stacktrace is below).
After falling, it restarts and tries to recover its state and restart Spark driver again. It happens over and over in an infinite cycle.
Namely, Spark tries to read DriverInfo state from zookeeper, but after reading it happens to be null in DriverInfo.worker.
https://issues.apache.org/jira/browse/SPARK-3150
Author: Tatiana Borisova <tanyatik@yandex.ru>
Closes#2062 from tanyatik/spark-3150 and squashes the following commits:
9936043 [Tatiana Borisova] Add initializing default values in DriverInfo.init()
In `SparkSubmitDriverBootstrapper`, we wait for the parent process to send us an `EOF` before finishing the application. This is applicable for the PySpark shell because we terminate the application the same way. However if we run a python application, for instance, the JVM actually never exits unless it receives a manual EOF from the user. This is causing a few tests to timeout.
We only need to do this for the PySpark shell because Spark submit runs as a python subprocess only in this case. Thus, the normal Spark shell doesn't need to go through this case even though it is also a REPL.
Thanks davies for reporting this.
Author: Andrew Or <andrewor14@gmail.com>
Closes#2170 from andrewor14/bootstrap-hotfix and squashes the following commits:
42963f5 [Andrew Or] Do not wait for EOF unless this is the pyspark shell
delete no used method in Utils
Author: scwf <wangfei1@huawei.com>
Closes#2160 from scwf/delete-no-use-method and squashes the following commits:
d8f6b0d [scwf] delete no use method in Utils
If we set both `spark.driver.extraClassPath` and `--driver-class-path`, then the latter correctly overrides the former. However, the value of the system property `spark.driver.extraClassPath` still uses the former, which is actually not added to the class path. This may cause some confusion...
Of course, this also affects other options (i.e. java options, library path, memory...).
Author: Andrew Or <andrewor14@gmail.com>
Closes#2154 from andrewor14/driver-submit-configs-fix and squashes the following commits:
17ec6fc [Andrew Or] Fix tests
0140836 [Andrew Or] Don't forget spark.driver.memory
e39d20f [Andrew Or] Also set spark.driver.extra* configs in client mode
compeleted stage only need to remove its own partitions that are no longer cached. However, "StorageTab" may lost some rdds which are cached actually. Not only in "StorageTab", "ExectutorTab" may also lose some rdd info which have been overwritten by last rdd in a same task.
1. "StorageTab": when multiple stages run simultaneously, completed stage will remove rdd info which belong to other stages that are still running.
2. "ExectutorTab": taskcontext may lose some "updatedBlocks" info of rdds in a dependency chain. Like the following example:
val r1 = sc.paralize(..).cache()
val r2 = r1.map(...).cache()
val n = r2.count()
When count the r2, r1 and r2 will be cached finally. So in CacheManager.getOrCompute, the taskcontext should contain "updatedBlocks" of r1 and r2. Currently, the "updatedBlocks" only contain the info of r2.
Author: uncleGen <hustyugm@gmail.com>
Closes#2131 from uncleGen/master_ui_fix and squashes the following commits:
a6a8a0b [uncleGen] fix some coding style
3a1bc15 [uncleGen] fix some error in unit test
56ea488 [uncleGen] there's some line too long
c82ba82 [uncleGen] Bug Fix: RDD info loss in "StorageTab" and "ExecutorTab"
As a workaround for SPARK-3015, the ContextCleaner was made "blocking", that is, it cleaned items one-by-one. But shuffles can take a long time to be deleted. Given that the RC for 1.1 is imminent, this PR makes a narrow change in the context cleaner - not wait for shuffle cleanups to complete. Also it changes the error messages on failure to delete to be milder warnings, as exceptions in the delete code path for one item does not really stop the actual functioning of the system.
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#2143 from tdas/cleaner-shuffle-fix and squashes the following commits:
9c84202 [Tathagata Das] Restoring default blocking behavior in ContextCleanerSuite, and added docs to identify that spark.cleaner.referenceTracking.blocking does not control shuffle.
2181329 [Tathagata Das] Mark shuffle cleanup as non-blocking.
e337cc2 [Tathagata Das] Changed semantics based on PR comments.
387b578 [Tathagata Das] Made ContextCleaner to not block on shuffles
This is an effort to bring the Windows scripts up to speed after recent splashing changes in #1845.
Author: Andrew Or <andrewor14@gmail.com>
Closes#2129 from andrewor14/windows-config and squashes the following commits:
881a8f0 [Andrew Or] Add reference to Windows taskkill
92e6047 [Andrew Or] Update a few comments (minor)
22b1acd [Andrew Or] Fix style again (minor)
afcffea [Andrew Or] Fix style (minor)
72004c2 [Andrew Or] Actually respect --driver-java-options
803218b [Andrew Or] Actually respect SPARK_*_CLASSPATH
eeb34a0 [Andrew Or] Update outdated comment (minor)
35caecc [Andrew Or] In Windows, actually kill Java processes on exit
f97daa2 [Andrew Or] Fix Windows spark shell stdin issue
83ebe60 [Andrew Or] Parse special driver configs in Windows (broken)
This is a HOTFIX for 1.1.
Author: Reynold Xin <rxin@apache.org>
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#2127 from rxin/SPARK-3224 and squashes the following commits:
effb1ce [Reynold Xin] Move log message.
49282b3 [Reynold Xin] Kay's feedback.
3f01847 [Reynold Xin] Merge pull request #2 from kayousterhout/SPARK-3224
796d282 [Kay Ousterhout] Added unit test for SPARK-3224
3d3d356 [Reynold Xin] Remove map output loc even for repeated FetchFaileds.
1dd3eb5 [Reynold Xin] [SPARK-3224] FetchFailed reduce stages should only show up once in the failed stages UI.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#1886 from sarutak/SPARK-2964 and squashes the following commits:
8ef8751 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2964
26e7c95 [Kousuke Saruta] Revert "Shorten timeout to more reasonable value"
ffb68fa [Kousuke Saruta] Modified spark-sql and start-thriftserver.sh to use bin/utils.sh
8c6f658 [Kousuke Saruta] Merge branch 'spark-3026' of https://github.com/liancheng/spark into SPARK-2964
81b43a8 [Cheng Lian] Shorten timeout to more reasonable value
a89e66d [Cheng Lian] Fixed command line options quotation in scripts
9c894d3 [Cheng Lian] Fixed bin/spark-sql -S option typo
be4736b [Cheng Lian] Report better error message when running JDBC/CLI without hive-thriftserver profile enabled
As of #1777 we log the name of the actor system when it binds to a port. The current name "spark" is super general and does not convey any meaning. For instance, the following line is taken from my driver log after setting `spark.driver.port` to 5001.
```
14/08/13 19:33:29 INFO Remoting: Remoting started; listening on addresses:
[akka.tcp://sparkandrews-mbp:5001]
14/08/13 19:33:29 INFO Remoting: Remoting now listens on addresses:
[akka.tcp://sparkandrews-mbp:5001]
14/08/06 13:40:05 INFO Utils: Successfully started service 'spark' on port 5001.
```
This commit renames this to "sparkDriver" and "sparkExecutor". The goal of this unambitious PR is simply to make the logged information more explicit without introducing any change in functionality.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1810 from andrewor14/service-name and squashes the following commits:
8c459ed [Andrew Or] Use a common variable for driver/executor actor system names
3a92843 [Andrew Or] Change actor name to sparkDriver and sparkExecutor
921363e [Andrew Or] Merge branch 'master' of github.com:apache/spark into service-name
c8c6a62 [Andrew Or] Do not include hyphens in actor name
1c1b42e [Andrew Or] Avoid spaces in akka system name
f644b55 [Andrew Or] Use more specific service name
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#1895 from sarutak/SPARK-2976 and squashes the following commits:
1cf7e69 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2976
d1e0666 [Kousuke Saruta] Modified styles
c5e80a4 [Kousuke Saruta] Remove tab from JavaPageRank.java and JavaKinesisWordCountASL.java
c003b36 [Kousuke Saruta] Removed tab from sorttable.js
rxin
Author: Xiangrui Meng <meng@databricks.com>
Closes#2120 from mengxr/sendMessageReliably and squashes the following commits:
b14400c [Xiangrui Meng] fix error message in sendMessageReliably
Just clean unused code which have been moved into ExternalSorter.
Author: Raymond Liu <raymond.liu@intel.com>
Closes#1882 from colorant/sortShuffleWriter and squashes the following commits:
e6337be [Raymond Liu] Clean unused code in SortShuffleWriter
RDD.countApprox(self, timeout, confidence=0.95)
:: Experimental ::
Approximate version of count() that returns a potentially incomplete
result within a timeout, even if not all tasks have finished.
>>> rdd = sc.parallelize(range(1000), 10)
>>> rdd.countApprox(1000, 1.0)
1000
RDD.sumApprox(self, timeout, confidence=0.95)
Approximate operation to return the sum within a timeout
or meet the confidence.
>>> rdd = sc.parallelize(range(1000), 10)
>>> r = sum(xrange(1000))
>>> (rdd.sumApprox(1000) - r) / r < 0.05
RDD.meanApprox(self, timeout, confidence=0.95)
:: Experimental ::
Approximate operation to return the mean within a timeout
or meet the confidence.
>>> rdd = sc.parallelize(range(1000), 10)
>>> r = sum(xrange(1000)) / 1000.0
>>> (rdd.meanApprox(1000) - r) / r < 0.05
True
Author: Davies Liu <davies.liu@gmail.com>
Closes#2095 from davies/approx and squashes the following commits:
e8c252b [Davies Liu] add approx API for RDD
Fix a typo in comment.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#2105 from viirya/fix_typo and squashes the following commits:
6596a80 [Liang-Chi Hsieh] fix typo.
In JVM 1.8.0, MaxPermSize is no longer supported.
In spark `stderr` output, there would be a line of
Java HotSpot(TM) 64-Bit Server VM warning: ignoring option MaxPermSize=128m; support was removed in 8.0
Author: Daoyuan Wang <daoyuan.wang@intel.com>
Closes#2011 from adrian-wang/maxpermsize and squashes the following commits:
ef1d660 [Daoyuan Wang] direct get java version in runtime
37db9c1 [Daoyuan Wang] code refine
3c1d554 [Daoyuan Wang] remove MaxPermSize option for jvm 1.8
Simple way to reproduce this in the UI:
```scala
val f = new java.io.File("/tmp/test")
f.delete()
sc.parallelize(1 to 2, 2).map(x => (x,x )).repartition(3).mapPartitionsWithContext { case (context, iter) =>
if (context.partitionId == 0) {
val f = new java.io.File("/tmp/test")
if (!f.exists) {
f.mkdir()
System.exit(0);
}
}
iter
}.count()
```
Author: Reynold Xin <rxin@apache.org>
Closes#1545 from rxin/stage-attempt and squashes the following commits:
3ee1d2a [Reynold Xin] - Rename attempt to retry in UI. - Properly report stage failure in FetchFailed.
40a6bd5 [Reynold Xin] Updated test suites.
c414c36 [Reynold Xin] Fixed the hanging in JobCancellationSuite.
b3e2eed [Reynold Xin] Oops previous code didn't compile.
0f36075 [Reynold Xin] Mark unknown stage attempt with id -1 and drop that in JobProgressListener.
6c08b07 [Reynold Xin] Addressed code review feedback.
4e5faa2 [Reynold Xin] [SPARK-2298] Encode stage attempt in SparkListener & UI.
In client deploy mode, the driver is launched from within `SparkSubmit`'s JVM. This means by the time we parse Spark configs from `spark-defaults.conf`, it is already too late to control certain properties of the driver's JVM. We currently ignore these configs in client mode altogether.
```
spark.driver.memory
spark.driver.extraJavaOptions
spark.driver.extraClassPath
spark.driver.extraLibraryPath
```
This PR handles these properties before launching the driver JVM. It achieves this by spawning a separate JVM that runs a new class called `SparkSubmitDriverBootstrapper`, which spawns `SparkSubmit` as a sub-process with the appropriate classpath, library paths, java opts and memory.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1845 from andrewor14/handle-configs-bash and squashes the following commits:
bed4bdf [Andrew Or] Change a few comments / messages (minor)
24dba60 [Andrew Or] Merge branch 'master' of github.com:apache/spark into handle-configs-bash
08fd788 [Andrew Or] Warn against external usages of SparkSubmitDriverBootstrapper
ff34728 [Andrew Or] Minor comments
51aeb01 [Andrew Or] Filter out JVM memory in Scala rather than Bash (minor)
9a778f6 [Andrew Or] Fix PySpark: actually kill driver on termination
d0f20db [Andrew Or] Don't pass empty library paths, classpath, java opts etc.
a78cb26 [Andrew Or] Revert a few changes in utils.sh (minor)
9ba37e2 [Andrew Or] Don't barf when the properties file does not exist
8867a09 [Andrew Or] A few more naming things (minor)
19464ad [Andrew Or] SPARK_SUBMIT_JAVA_OPTS -> SPARK_SUBMIT_OPTS
d6488f9 [Andrew Or] Merge branch 'master' of github.com:apache/spark into handle-configs-bash
1ea6bbe [Andrew Or] SparkClassLauncher -> SparkSubmitDriverBootstrapper
a91ea19 [Andrew Or] Fix precedence of library paths, classpath, java opts and memory
158f813 [Andrew Or] Remove "client mode" boolean argument
c84f5c8 [Andrew Or] Remove debug print statement (minor)
b71f52b [Andrew Or] Revert a few more changes (minor)
7d94a8d [Andrew Or] Merge branch 'master' of github.com:apache/spark into handle-configs-bash
3a8235d [Andrew Or] Only parse the properties file if special configs exist
c37e08d [Andrew Or] Revert a few more changes
a396eda [Andrew Or] Nullify my own hard work to simplify bash
0effa1e [Andrew Or] Add code in Scala that handles special configs
c886568 [Andrew Or] Fix lines too long + a few comments / style (minor)
7a4190a [Andrew Or] Merge branch 'master' of github.com:apache/spark into handle-configs-bash
7396be2 [Andrew Or] Explicitly comment that multi-line properties are not supported
fa11ef8 [Andrew Or] Parse the properties file only if the special configs exist
371cac4 [Andrew Or] Add function prefix (minor)
be99eb3 [Andrew Or] Fix tests to not include multi-line configs
bd0d468 [Andrew Or] Simplify parsing config file by ignoring multi-line arguments
56ac247 [Andrew Or] Use eval and set to simplify splitting
8d4614c [Andrew Or] Merge branch 'master' of github.com:apache/spark into handle-configs-bash
aeb79c7 [Andrew Or] Merge branch 'master' of github.com:apache/spark into handle-configs-bash
2732ac0 [Andrew Or] Integrate BASH tests into dev/run-tests + log error properly
8d26a5c [Andrew Or] Add tests for bash/utils.sh
4ae24c3 [Andrew Or] Fix bug: escape properly in quote_java_property
b3c4cd5 [Andrew Or] Fix bug: count the number of quotes instead of detecting presence
c2273fc [Andrew Or] Fix typo (minor)
e793e5f [Andrew Or] Handle multi-line arguments
5d8f8c4 [Andrew Or] Merge branch 'master' of github.com:apache/spark into submit-driver-extra
c7b9926 [Andrew Or] Minor changes to spark-defaults.conf.template
a992ae2 [Andrew Or] Escape spark.*.extraJavaOptions correctly
aabfc7e [Andrew Or] escape -> split (minor)
45a1eb9 [Andrew Or] Fix bug: escape escaped backslashes and quotes properly...
1cdc6b1 [Andrew Or] Fix bug: escape escaped double quotes properly
c854859 [Andrew Or] Add small comment
c13a2cb [Andrew Or] Merge branch 'master' of github.com:apache/spark into submit-driver-extra
8e552b7 [Andrew Or] Include an example of spark.*.extraJavaOptions
de765c9 [Andrew Or] Print spark-class command properly
a4df3c4 [Andrew Or] Move parsing and escaping logic to utils.sh
dec2343 [Andrew Or] Only export variables if they exist
fa2136e [Andrew Or] Escape Java options + parse java properties files properly
ef12f74 [Andrew Or] Minor formatting
4ec22a1 [Andrew Or] Merge branch 'master' of github.com:apache/spark into submit-driver-extra
e5cfb46 [Andrew Or] Collapse duplicate code + fix potential whitespace issues
4edcaa8 [Andrew Or] Redirect stdout to stderr for python
130f295 [Andrew Or] Handle spark.driver.memory too
98dd8e3 [Andrew Or] Add warning if properties file does not exist
8843562 [Andrew Or] Fix compilation issues...
75ee6b4 [Andrew Or] Remove accidentally added file
63ed2e9 [Andrew Or] Merge branch 'master' of github.com:apache/spark into submit-driver-extra
0025474 [Andrew Or] Revert SparkSubmit handling of --driver-* options for only cluster mode
a2ab1b0 [Andrew Or] Parse spark.driver.extra* in bash
250cb95 [Andrew Or] Do not ignore spark.driver.extra* for client mode
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2060 from sarutak/SPARK-3149 and squashes the following commits:
1cc89af [Kousuke Saruta] Modified log message of accepting connection
#1891 was to avoid IOException when EventLogging is enabled.
The solution used ShutdownHookManager but it was defined only Hadoop 2.x. Hadoop 1.x don't have ShutdownHookManager so #1891 doesn't compile on Hadoop 1.x
Now, I had a compromised solution for both Hadoop 1.x and 2.x.
Only for FileLogger, an unique FileSystem object is created.
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#1970 from sarutak/SPARK-2970 and squashes the following commits:
240c91e [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2970
0e7b45d [Kousuke Saruta] Revert "[SPARK-2970] [SQL] spark-sql script ends with IOException when EventLogging is enabled"
e1262ec [Kousuke Saruta] Modified Filelogger to use unique FileSystem instance
This PR fixes two bugs related to `spark.local.dirs` and `SPARK_LOCAL_DIRS`, one where `Utils.getLocalDir()` might return an invalid directory (SPARK-2974) and another where the `SPARK_LOCAL_DIRS` override didn't affect the driver, which could cause problems when running tasks in local mode (SPARK-2975).
This patch fixes both issues: the new `Utils.getOrCreateLocalRootDirs(conf: SparkConf)` utility method manages the creation of local directories and handles the precedence among the different configuration options, so we should see the same behavior whether we're running in local mode or on a worker.
It's kind of a pain to mock out environment variables in tests (no easy way to mock System.getenv), so I added a `private[spark]` method to SparkConf for accessing environment variables (by default, it just delegates to System.getenv). By subclassing SparkConf and overriding this method, we can mock out SPARK_LOCAL_DIRS in tests.
I also fixed a typo in PySpark where we used `SPARK_LOCAL_DIR` instead of `SPARK_LOCAL_DIRS` (I think this was technically innocuous, but it seemed worth fixing).
Author: Josh Rosen <joshrosen@apache.org>
Closes#2002 from JoshRosen/local-dirs and squashes the following commits:
efad8c6 [Josh Rosen] Address review comments:
1dec709 [Josh Rosen] Minor updates to Javadocs.
7f36999 [Josh Rosen] Use env vars to detect if running in YARN container.
399ac25 [Josh Rosen] Update getLocalDir() documentation.
bb3ad89 [Josh Rosen] Remove duplicated YARN getLocalDirs() code.
3e92d44 [Josh Rosen] Move local dirs override logic into Utils; fix bugs:
b2c4736 [Josh Rosen] Add failing tests for SPARK-2974 and SPARK-2975.
007298b [Josh Rosen] Allow environment variables to be mocked in tests.
6d9259b [Josh Rosen] Fix typo in PySpark: SPARK_LOCAL_DIR should be SPARK_LOCAL_DIRS
This is a re-implementation of TorrentBroadcast, with the following changes:
1. Removes most of the mutable, transient state from TorrentBroadcast (e.g. totalBytes, num of blocks fetched).
2. Removes TorrentInfo and TorrentBlock
3. Replaces the BlockManager.getSingle call in readObject with a getLocal, resuling in one less RPC call to the BlockManagerMasterActor to find the location of the block.
4. Removes the metadata block, resulting in one less block to fetch.
5. Removes an extra memory copy for deserialization (by using Java's SequenceInputStream).
Basically for a regular broadcasted object with only one block, the number of RPC calls goes from 5+1 to 2+1).
Old TorrentBroadcast for object of a single block:
1 RPC to ask for location of the broadcast variable
1 RPC to ask for location of the metadata block
1 RPC to fetch the metadata block
1 RPC to ask for location of the first data block
1 RPC to fetch the first data block
1 RPC to tell the driver we put the first data block in
i.e. 5 + 1
New TorrentBroadcast for object of a single block:
1 RPC to ask for location of the first data block
1 RPC to get the first data block
1 RPC to tell the driver we put the first data block in
i.e. 2 + 1
Author: Reynold Xin <rxin@apache.org>
Closes#2030 from rxin/torrentBroadcast and squashes the following commits:
5bacb9d [Reynold Xin] Always add the object to driver's block manager.
0d8ed5b [Reynold Xin] Added getBytes to BlockManager and uses that in TorrentBroadcast.
2d6a5fb [Reynold Xin] Use putBytes/getRemoteBytes throughout.
3670f00 [Reynold Xin] Code review feedback.
c1185cd [Reynold Xin] [SPARK-3119] Re-implementation of TorrentBroadcast.
Previous pull request (#1907) was reverted. This brings it back. Still looking into the hang.
Author: Reynold Xin <rxin@apache.org>
Closes#1971 from rxin/netty1 and squashes the following commits:
b0be96f [Reynold Xin] Added test to make sure outstandingRequests are cleaned after firing the events.
4c6d0ee [Reynold Xin] Pass callbacks cleanly.
603dce7 [Reynold Xin] Upgrade Netty to 4.0.23 to fix the DefaultFileRegion bug.
88be1d4 [Reynold Xin] Downgrade to 4.0.21 to work around a bug in writing DefaultFileRegion.
002626a [Reynold Xin] Remove netty-test-file.txt.
db6e6e0 [Reynold Xin] Revert "Revert "[SPARK-2468] Netty based block server / client module""
Move a bracket in validateSettings of SparkConf
Author: hzw19900416 <carlmartinmax@gmail.com>
Closes#2012 from hzw19900416/codereading and squashes the following commits:
e717fb6 [hzw19900416] Move a bracket in validateSettings of SparkConf
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#2000 from sarutak/SPARK-3089 and squashes the following commits:
02dfdea [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-3089
e759ce7 [Kousuke Saruta] Improved error message when closing SendingConnection
Author: Reynold Xin <rxin@apache.org>
Closes#2028 from rxin/torrentBroadcast and squashes the following commits:
92c62a5 [Reynold Xin] Revert the MEMORY_AND_DISK_SER changes.
03a5221 [Reynold Xin] [SPARK-3116] Remove the excessive lockings in TorrentBroadcast
Due to the way Yarn runs things through bash, normal quoting doesn't
work as expected. This change applies the necessary voodoo to the user
args to avoid issues with bash and special characters.
The change also uncovered an issue with the event logger app name
sanitizing code; it wasn't cleaning up all "bad" characters, so
sometimes it would fail to create the log dirs. I just added some
more bad character replacements.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#1724 from vanzin/SPARK-2718 and squashes the following commits:
cc84b89 [Marcelo Vanzin] Review feedback.
c1a257a [Marcelo Vanzin] Add test for backslashes.
55571d4 [Marcelo Vanzin] Unbreak yarn-client.
515613d [Marcelo Vanzin] [SPARK-2718] [yarn] Handle quotes and other characters in user args.
Instead of keeping copies in all pages, just reference the values
kept in the base SparkUI instance (by making them available via
getters).
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#1252 from vanzin/SPARK-2169 and squashes the following commits:
4412fc6 [Marcelo Vanzin] Simplify UIUtils.headerSparkPage signature.
4e5d35a [Marcelo Vanzin] [SPARK-2169] Don't copy appName / basePath everywhere.
Author: Chandan Kumar <chandan.kumar@imaginea.com>
Closes#1787 from nrchandan/spark-2862 and squashes the following commits:
a76bbf6 [Chandan Kumar] [SPARK-2862] Fix for a broken test case and add new test cases
4211eea [Chandan Kumar] [SPARK-2862] Add Scala bug id
13854f1 [Chandan Kumar] [SPARK-2862] Use shorthand range notation to avoid Scala bug
there's no need to use masterLock in Worker now since all communications are within Akka actor
Author: CrazyJvm <crazyjvm@gmail.com>
Closes#2008 from CrazyJvm/no-need-master-lock and squashes the following commits:
dd39e20 [CrazyJvm] fix format
58e7fa5 [CrazyJvm] there's no need to use masterLock now since all communications are within Akka actor
cc JoshRosen sarutak
Author: GuoQiang Li <witgo@qq.com>
Closes#1989 from witgo/cancel_ackTimeoutMonitor and squashes the following commits:
4a700fa [GuoQiang Li] In the stop method of ConnectionManager to cancel the ackTimeoutMonitor
Passing large object by py4j is very slow (cost much memory), so pass broadcast objects via files (similar to parallelize()).
Add an option to keep object in driver (it's False by default) to save memory in driver.
Author: Davies Liu <davies.liu@gmail.com>
Closes#1912 from davies/broadcast and squashes the following commits:
e06df4a [Davies Liu] load broadcast from disk in driver automatically
db3f232 [Davies Liu] fix serialization of accumulator
631a827 [Davies Liu] Merge branch 'master' into broadcast
c7baa8c [Davies Liu] compress serrialized broadcast and command
9a7161f [Davies Liu] fix doc tests
e93cf4b [Davies Liu] address comments: add test
6226189 [Davies Liu] improve large broadcast
This is intended to fix SPARK-2977. Before, there was an implicit ordering dependency where we needed to know the ShuffleManager implementation before creating the ShuffleBlockManager. This patch makes that dependency explicit by adding ShuffleManager to a bunch of constructors.
I think it's a little odd for BlockManager to take a ShuffleManager only to pass it to ShuffleBlockManager without using it itself; there's an opportunity to clean this up later if we sever the circular dependencies between BlockManager and other components and pass those components to BlockManager's constructor.
Author: Josh Rosen <joshrosen@apache.org>
Closes#1976 from JoshRosen/SPARK-2977 and squashes the following commits:
a9cd1e1 [Josh Rosen] [SPARK-2977] Ensure ShuffleManager is created before ShuffleBlockManager.
Author: Reynold Xin <rxin@apache.org>
Closes#1948 from rxin/kryo and squashes the following commits:
a3a80d8 [Reynold Xin] [SPARK-3046] use executor's class loader as the default serializer classloader
3d13277 [Reynold Xin] Reverted that in TestJavaSerializerImpl too.
196f3dc [Reynold Xin] Ok one more commit to revert the classloader change.
c49b50c [Reynold Xin] Removed JavaSerializer change.
afbf37d [Reynold Xin] Moved the test case also.
a2e693e [Reynold Xin] Removed the Kryo bug fix from this pull request.
c81bd6c [Reynold Xin] Use defaultClassLoader when executing user specified custom registrator.
68f261e [Reynold Xin] Added license check excludes.
0c28179 [Reynold Xin] [SPARK-3045] Make Serializer interface Java friendly [SPARK-3046] Set executor's class loader as the default serializer class loader
More detail on the issue is described in [SPARK-3015](https://issues.apache.org/jira/browse/SPARK-3015), but the TLDR is if we send too many blocking Akka messages that are dependent on each other in quick successions, then we end up causing a few of these messages to time out and ultimately kill the executors. As of #1498, we broadcast each RDD whether or not it is persisted. This means if we create many RDDs (each of which becomes a broadcast) and the driver performs a GC that cleans up all of these broadcast blocks, then we end up sending many `RemoveBroadcast` messages in parallel and trigger the chain of blocking messages at high frequencies.
We do not know of the Akka-level root cause yet, so this is intended to be a temporary solution until we identify the real issue. I have done some preliminary testing of enabling blocking and observed that the queue length remains quite low (< 1000) even under very intensive workloads.
In the long run, we should do something more sophisticated to allow a limited degree of parallelism through batching clean up tasks or processing them in a sliding window. In the longer run, we should clean up the whole `BlockManager*` message passing interface to avoid unnecessarily awaiting on futures created from Akka asks.
tdas pwendell mengxr
Author: Andrew Or <andrewor14@gmail.com>
Closes#1931 from andrewor14/reference-blocking and squashes the following commits:
d0f7195 [Andrew Or] Merge branch 'master' of github.com:apache/spark into reference-blocking
ce9daf5 [Andrew Or] Remove logic for logging queue length
111192a [Andrew Or] Add missing space in log message (minor)
a183b83 [Andrew Or] Switch order of code blocks (minor)
9fd1fe6 [Andrew Or] Remove outdated log
104b366 [Andrew Or] Use the actual reference queue length
0b7e768 [Andrew Or] Block on cleaning tasks by default + log error on queue full
The serializer is not always used in an executor thread (e.g. connection manager, broadcast), in which case the classloader might not have the user jar set, leading to corruption in deserialization.
https://issues.apache.org/jira/browse/SPARK-3046https://issues.apache.org/jira/browse/SPARK-2878
Author: Reynold Xin <rxin@apache.org>
Closes#1972 from rxin/kryoBug and squashes the following commits:
c1c7bf0 [Reynold Xin] Made change to JavaSerializer.
7204c33 [Reynold Xin] Added imports back.
d879e67 [Reynold Xin] [SPARK-3046] use executor's class loader as the default serializer class loader.
...Update
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1961 from sryza/sandy-spark-3028 and squashes the following commits:
dccdff5 [Sandy Ryza] Fix compile error
f883ded [Sandy Ryza] SPARK-3028. sparkEventToJson should support SparkListenerExecutorMetricsUpdate
Not supported in Scala 2.11. Split them into separate methods instead.
Author: Anand Avati <avati@redhat.com>
Closes#1704 from avati/SPARK-1812-default-args and squashes the following commits:
3e3924a [Anand Avati] SPARK-1812: Add Mima excludes for the broken ABI
901dfc7 [Anand Avati] SPARK-1812: core - Fix overloaded methods with default arguments
07f00af [Anand Avati] SPARK-1812: streaming - Fix overloaded methods with default arguments
I find it useful to see where in an RDD's DAG data is cached, so I figured others might too.
I've added both the caching level, and the actual memory state of the RDD.
Some of this is redundant with the web UI (notably the actual memory state), but (a) that is temporary, and (b) putting it in the DAG tree shows some context that can help a lot.
For example:
```
(4) ShuffledRDD[3] at reduceByKey at <console>:14
+-(4) MappedRDD[2] at map at <console>:14
| MapPartitionsRDD[1] at mapPartitions at <console>:12
| ParallelCollectionRDD[0] at parallelize at <console>:12
```
should change to
```
(4) ShuffledRDD[3] at reduceByKey at <console>:14 [Memory Deserialized 1x Replicated]
| CachedPartitions: 4; MemorySize: 50.8 MB; TachyonSize: 0.0 B; DiskSize: 0.0 B
+-(4) MappedRDD[2] at map at <console>:14 [Memory Deserialized 1x Replicated]
| MapPartitionsRDD[1] at mapPartitions at <console>:12 [Memory Deserialized 1x Replicated]
| CachedPartitions: 4; MemorySize: 109.1 MB; TachyonSize: 0.0 B; DiskSize: 0.0 B
| ParallelCollectionRDD[0] at parallelize at <console>:12 [Memory Deserialized 1x Replicated]
```
Author: Nathan Kronenfeld <nkronenfeld@oculusinfo.com>
Closes#1535 from nkronenfeld/feature/debug-caching2 and squashes the following commits:
40490bc [Nathan Kronenfeld] Back out DeveloperAPI and arguments to RDD.toDebugString, reinstate memory output
794e6a3 [Nathan Kronenfeld] Attempt to merge mima changes from master
6fe9e80 [Nathan Kronenfeld] Add exclusions to allow for signature change in toDebugString (will back out if necessary)
31d6769 [Nathan Kronenfeld] Attempt to get rid of style errors. Add comments for the new memory usage parameter.
a0f6f76 [Nathan Kronenfeld] Add parameter to RDD.toDebugString to allow detailed memory info to be shown or not. Default is for it not to be shown.
f8f565a [Nathan Kronenfeld] Fix code style error
8f54287 [Nathan Kronenfeld] Changed string addition to string interpolation as per PR comments
2a0cd4d [Nathan Kronenfeld] Fixed a small formatting issue I forgot to copy over from the old branch
8fbecb6 [Nathan Kronenfeld] Add caching information to rdd.toDebugString
JIRA: https://issues.apache.org/jira/browse/SPARK-2736
This patch includes:
1. An Avro converter that converts Avro data types to Python. It handles all 3 Avro data mappings (Generic, Specific and Reflect).
2. An example Python script for reading Avro files using AvroKeyInputFormat and the converter.
3. Fixing a classloading issue.
cc @MLnick @JoshRosen @mateiz
Author: Kan Zhang <kzhang@apache.org>
Closes#1916 from kanzhang/SPARK-2736 and squashes the following commits:
02443f8 [Kan Zhang] [SPARK-2736] Adding .avsc files to .rat-excludes
f74e9a9 [Kan Zhang] [SPARK-2736] nit: clazz -> className
82cc505 [Kan Zhang] [SPARK-2736] Update data sample
0be7761 [Kan Zhang] [SPARK-2736] Example pyspark script and data files
c8e5881 [Kan Zhang] [SPARK-2736] Trying to work with all 3 Avro data models
2271a5b [Kan Zhang] [SPARK-2736] Using the right class loader to find Avro classes
536876b [Kan Zhang] [SPARK-2736] Adding Avro to Java converter
This is a rewrite of the original Netty module that was added about 1.5 years ago. The old code was turned off by default and didn't really work because it lacked a frame decoder (only worked with very very small blocks).
For this pull request, I tried to make the changes non-instrusive to the rest of Spark. I only added an init and shutdown to BlockManager/DiskBlockManager, and a bunch of comments to help me understand the existing code base.
Compared with the old Netty module, this one features:
- It appears to work :)
- SPARK-2941: option to specicy nio vs oio vs epoll for channel/transport. By default nio is used. (Not using Epoll yet because I have found some bugs with its implementation)
- SPARK-2943: options to specify send buf and receive buf for users who want to do hyper tuning
- SPARK-2942: io errors are reported from server to client (the protocol uses negative length to indicate error)
- SPARK-2940: fetching multiple blocks in a single request to reduce syscalls
- SPARK-2959: clients share a single thread pool
- SPARK-2990: use PooledByteBufAllocator to reduce GC (basically a Netty managed pool of buffers with jmalloc)
- SPARK-2625: added fetchWaitTime metric and fixed thread-safety issue in metrics update.
- SPARK-2367: bump Netty version to 4.0.21.Final to address an Epoll bug (https://groups.google.com/forum/#!topic/netty/O7m-HxCJpCA)
Compared with the existing communication manager, this one features:
- IMO it is substantially easier to understand
- zero-copy send for the server for on-disk blocks
- one-copy receive (due to a frame decoder)
- don't quote me on this, but I think a lot less sys calls
- SPARK-2990: use PooledByteBufAllocator to reduce GC (basically a Netty managed pool of buffers with jmalloc)
- SPARK-2941: option to specicy nio vs oio vs epoll for channel/transport. By default nio is used. (Not using Epoll yet because I have found some bugs with its implementation)
- SPARK-2943: options to specify send buf and receive buf for users who want to do hyper tuning
TODOs before it can fully replace the existing ConnectionManager, if that ever happens (most of them should probably be done in separate PRs since this needs to be turned on explicitly)
- [x] Basic test cases
- [ ] More unit/integration tests for failures
- [ ] Performance analysis
- [ ] Support client connection reuse so we don't need to keep opening new connections (not sure how useful this would be)
- [ ] Support putting blocks in addition to fetching blocks (i.e. two way transfer)
- [x] Support serving non-disk blocks
- [ ] Support SASL authentication
For a more comprehensive list, see https://issues.apache.org/jira/browse/SPARK-2468
Thanks to @coderplay for peer coding with me on a Sunday.
Author: Reynold Xin <rxin@apache.org>
Closes#1907 from rxin/netty and squashes the following commits:
f921421 [Reynold Xin] Upgrade Netty to 4.0.22.Final to fix another Epoll bug.
4b174ca [Reynold Xin] Shivaram's code review comment.
4a3dfe7 [Reynold Xin] Switched to nio for default (instead of epoll on Linux).
56bfb9d [Reynold Xin] Bump Netty version to 4.0.21.Final for some bug fixes.
b443a4b [Reynold Xin] Added debug message to help debug Jenkins failures.
57fc4d7 [Reynold Xin] Added test cases for BlockHeaderEncoder and BlockFetchingClientHandlerSuite.
22623e9 [Reynold Xin] Added exception handling and test case for BlockServerHandler and BlockFetchingClientHandler.
6550dd7 [Reynold Xin] Fixed block mgr init bug.
60c2edf [Reynold Xin] Beefed up server/client integration tests.
38d88d5 [Reynold Xin] Added missing test files.
6ce3f3c [Reynold Xin] Added some basic test cases.
47f7ce0 [Reynold Xin] Created server and client packages and moved files there.
b16f412 [Reynold Xin] Added commit count.
f13022d [Reynold Xin] Remove unused clone() in BlockFetcherIterator.
c57d68c [Reynold Xin] Added back missing files.
842dfa7 [Reynold Xin] Made everything work with proper reference counting.
3fae001 [Reynold Xin] Connected the new netty network module with rest of Spark.
1a8f6d4 [Reynold Xin] Completed protocol documentation.
2951478 [Reynold Xin] New Netty implementation.
cc7843d [Reynold Xin] Basic skeleton.
Note this also passes the TaskContext itself to the TaskCompletionListener. In the future we can mark TaskContext with the exception object if exception occurs during task execution.
Author: Reynold Xin <rxin@apache.org>
Closes#1938 from rxin/TaskContext and squashes the following commits:
145de43 [Reynold Xin] Added JavaTaskCompletionListenerImpl for Java API friendly guarantee.
f435ea5 [Reynold Xin] Added license header for TaskCompletionListener.
dc4ed27 [Reynold Xin] [SPARK-3027] TaskContext: tighten the visibility and provide Java friendly callback API
...ationInfo is initialized properly after deserialization
Author: Jacek Lewandowski <lewandowski.jacek@gmail.com>
Closes#1947 from jacek-lewandowski/master and squashes the following commits:
713b2f1 [Jacek Lewandowski] SPARK-3009: Reverted readObject method in ApplicationInfo so that ApplicationInfo is initialized properly after deserialization
- Added override.
- Marked some variables as private.
Author: Reynold Xin <rxin@apache.org>
Closes#1943 from rxin/metricsSource and squashes the following commits:
fbfa943 [Reynold Xin] Minor cleanup of metrics.Source. - Added override. - Marked some variables as private.
The previous behaviour of swallowing ClassNotFound exceptions when running a custom Kryo registrator could lead to difficult to debug problems later on at serialisation / deserialisation time, see SPARK-2878. Instead it is better to fail fast.
Added test case.
Author: Graham Dennis <graham.dennis@gmail.com>
Closes#1827 from GrahamDennis/feature/spark-2893 and squashes the following commits:
fbe4cb6 [Graham Dennis] [SPARK-2878]: Update the test case to match the updated exception message
65e53c5 [Graham Dennis] [SPARK-2893]: Improve message when a spark.kryo.registrator fails.
f480d85 [Graham Dennis] [SPARK-2893] Fix typo.
b59d2c2 [Graham Dennis] SPARK-2893: Do not swallow Exceptions when running a custom spark.kryo.registrator
Currently, local execution of Spark jobs is only used by take(), and it can be problematic as it can load a significant amount of data onto the driver. The worst case scenarios occur if the RDD is cached (guaranteed to load whole partition), has very large elements, or the partition is just large and we apply a filter with high selectivity or computational overhead.
Additionally, jobs that run locally in this manner do not show up in the web UI, and are thus harder to track or understand what is occurring.
This PR adds a flag to disable local execution, which is turned OFF by default, with the intention of perhaps eventually removing this functionality altogether. Removing it now is a tougher proposition since it is part of the public runJob API. An alternative solution would be to limit the flag to take()/first() to avoid impacting any external users of this API, but such usage (or, at least, reliance upon the feature) is hopefully minimal.
Author: Aaron Davidson <aaron@databricks.com>
Closes#1321 from aarondav/allowlocal and squashes the following commits:
136b253 [Aaron Davidson] Fix DAGSchedulerSuite
5599d55 [Aaron Davidson] [RFC] Disable local execution of Spark jobs by default
Author: Patrick Wendell <pwendell@gmail.com>
Closes#1933 from pwendell/speculation and squashes the following commits:
33a3473 [Patrick Wendell] Use OpenHashSet
8ce2ff0 [Patrick Wendell] SPARK-3020: Print completed indices rather than tasks in web UI
Author: Zhang, Liye <liye.zhang@intel.com>
Closes#1892 from liyezhang556520/lazy_memory_request and squashes the following commits:
335ab61 [Zhang, Liye] [SPARK-1777 (partial)] bugfix: make size of requested memory correctly
Since this is a file to file copy, using transferTo should be faster.
Author: Raymond Liu <raymond.liu@intel.com>
Closes#1884 from colorant/externalSorter and squashes the following commits:
6e42f3c [Raymond Liu] More code into copyStream
bfb496b [Raymond Liu] Use transferTo when copy merge files in ExternalSorter
Instead of requiring "org.apache.spark.io.LZ4CompressionCodec", it is easier for users if Spark just accepts "lz4", "lzf", "snappy".
Author: Reynold Xin <rxin@apache.org>
Closes#1873 from rxin/compressionCodecShortForm and squashes the following commits:
9f50962 [Reynold Xin] Specify short-form compression codec names first.
63f78ee [Reynold Xin] Updated configuration documentation.
47b3848 [Reynold Xin] [SPARK-2953] Allow using short names for io compression codecs
This addresses SPARK-2931, a bug where getAllowedLocalityLevel() could throw ArrayIndexOutOfBoundsException. The fix here is to reset currentLocalityIndex after recomputing the locality levels.
Thanks to kayousterhout, mridulm, and lirui-intel for helping me to debug this.
Author: Josh Rosen <joshrosen@apache.org>
Closes#1896 from JoshRosen/SPARK-2931 and squashes the following commits:
48b60b5 [Josh Rosen] Move FakeRackUtil.cleanUp() info beforeEach().
6fec474 [Josh Rosen] Set currentLocalityIndex after recomputing locality levels.
9384897 [Josh Rosen] Update SPARK-2931 test to reflect changes in 63bdb1f41b.
9ecd455 [Josh Rosen] Apply @mridulm's patch for reproducing SPARK-2931.
The Netty network module was originally written when Scala 2.9.x had a bug that prevents a pure Scala implementation, and a subset of the files were done in Java. We have since upgraded to Scala 2.10, and can migrate all Java files now to Scala.
https://github.com/netty/netty/issues/781https://github.com/mesos/spark/pull/522
Author: Reynold Xin <rxin@apache.org>
Closes#1865 from rxin/netty and squashes the following commits:
332422f [Reynold Xin] Code review feedback
ca9eeee [Reynold Xin] Minor update.
7f1434b [Reynold Xin] [SPARK-2936] Migrate Netty network module from Java to Scala
To enable Python consistency and `Experimental` label of the `sampleByKeyExact` API.
Author: Doris Xin <doris.s.xin@gmail.com>
Author: Xiangrui Meng <meng@databricks.com>
Closes#1866 from dorx/stratified and squashes the following commits:
0ad97b2 [Doris Xin] reviewer comments.
2948aae [Doris Xin] remove unrelated changes
e990325 [Doris Xin] Merge branch 'master' into stratified
555a3f9 [Doris Xin] separate out sampleByKeyExact as its own API
616e55c [Doris Xin] merge master
245439e [Doris Xin] moved minSamplingRate to getUpperBound
eaf5771 [Doris Xin] bug fixes.
17a381b [Doris Xin] fixed a merge issue and a failed unit
ea7d27f [Doris Xin] merge master
b223529 [Xiangrui Meng] use approx bounds for poisson fix poisson mean for waitlisting add unit tests for Java
b3013a4 [Xiangrui Meng] move math3 back to test scope
eecee5f [Doris Xin] Merge branch 'master' into stratified
f4c21f3 [Doris Xin] Reviewer comments
a10e68d [Doris Xin] style fix
a2bf756 [Doris Xin] Merge branch 'master' into stratified
680b677 [Doris Xin] use mapPartitionWithIndex instead
9884a9f [Doris Xin] style fix
bbfb8c9 [Doris Xin] Merge branch 'master' into stratified
ee9d260 [Doris Xin] addressed reviewer comments
6b5b10b [Doris Xin] Merge branch 'master' into stratified
254e03c [Doris Xin] minor fixes and Java API.
4ad516b [Doris Xin] remove unused imports from PairRDDFunctions
bd9dc6e [Doris Xin] unit bug and style violation fixed
1fe1cff [Doris Xin] Changed fractionByKey to a map to enable arg check
944a10c [Doris Xin] [SPARK-2145] Add lower bound on sampling rate
0214a76 [Doris Xin] cleanUp
90d94c0 [Doris Xin] merge master
9e74ab5 [Doris Xin] Separated out most of the logic in sampleByKey
7327611 [Doris Xin] merge master
50581fc [Doris Xin] added a TODO for logging in python
46f6c8c [Doris Xin] fixed the NPE caused by closures being cleaned before being passed into the aggregate function
7e1a481 [Doris Xin] changed the permission on SamplingUtil
1d413ce [Doris Xin] fixed checkstyle issues
9ee94ee [Doris Xin] [SPARK-2082] stratified sampling in PairRDDFunctions that guarantees exact sample size
e3fd6a6 [Doris Xin] Merge branch 'master' into takeSample
7cab53a [Doris Xin] fixed import bug in rdd.py
ffea61a [Doris Xin] SPARK-1939: Refactor takeSample method in RDD
1441977 [Doris Xin] SPARK-1939 Refactor takeSample method in RDD to use ScaSRS
1. do not use signal handler for SIGCHILD, it's easy to cause deadlock
2. handle EINTR during accept()
3. pass errno into JVM
4. handle EAGAIN during fork()
Now, it can pass 50k tasks tests in 180 seconds.
Author: Davies Liu <davies.liu@gmail.com>
Closes#1842 from davies/qa and squashes the following commits:
f0ea451 [Davies Liu] fix lint
03a2e8c [Davies Liu] cleanup dead children every seconds
32cb829 [Davies Liu] fix lint
0cd0817 [Davies Liu] fix bugs in deamon.py
The JobLogger is very useful for performing offline performance profiling of Spark jobs. GC Time and Shuffle Write time are available in TaskMetrics but are currently missed from the JobLogger output. This patch adds these two fields.
~~Since this is a small change, I didn't create a JIRA. Let me know if I should do that.~~
cc kayousterhout
Author: Shivaram Venkataraman <shivaram@cs.berkeley.edu>
Closes#1869 from shivaram/job-logger and squashes the following commits:
1b709fc [Shivaram Venkataraman] Add a space before GC_TIME
c418105 [Shivaram Venkataraman] Add gc time and shuffle write time to JobLogger
Author: GuoQiang Li <witgo@qq.com>
Closes#1876 from witgo/remove_semicolon_in_Task_scala and squashes the following commits:
c6ea732 [GuoQiang Li] Remove extra semicolon in Task.scala
This helps us log UpdateBlockInfo properly once #1870 is merged.
Author: Reynold Xin <rxin@apache.org>
Closes#1872 from rxin/UpdateBlockInfo and squashes the following commits:
0cee1c2 [Reynold Xin] Turn UpdateBlockInfo into case class.
As sryza reported, spark-shell doesn't accept any flags.
The root cause is wrong usage of spark-submit in spark-shell and it come to the surface by #1801
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes#1715, Closes#1864, and Closes#1861Closes#1825 from sarutak/SPARK-2894 and squashes the following commits:
47f3510 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2894
2c899ed [Kousuke Saruta] Removed useless code from java_gateway.py
98287ed [Kousuke Saruta] Removed useless code from java_gateway.py
513ad2e [Kousuke Saruta] Modified util.sh to enable to use option including white spaces
28a374e [Kousuke Saruta] Modified java_gateway.py to recognize arguments
5afc584 [Cheng Lian] Filter out spark-submit options when starting Python gateway
e630d19 [Cheng Lian] Fixing pyspark and spark-shell CLI options
Pedantry is underrated
Author: Chris Cope <ccope@resilientscience.com>
Closes#1859 from copester/master and squashes the following commits:
0fb4499 [Chris Cope] [SPARK-1766] sorted functions to meet pedantic requirements
Tested and ready to merge.
Author: Chandan Kumar <chandan.kumar@imaginea.com>
Closes#1786 from nrchandan/spark-2861 and squashes the following commits:
cb0bc1e [Chandan Kumar] [SPARK-2861] Fix a typo in the histogram doc comment
6a2a71b [Chandan Kumar] SPARK-2861. Fix Doc comment of histogram method
In SPARK-1946(PR #900), configuration <code>spark.scheduler.minRegisteredExecutorsRatio</code> was introduced. However, in standalone mode, there is a race condition where isReady() can return true because totalExpectedExecutors has not been correctly set.
Because expected executors is uncertain in standalone mode, the PR try to use CPU cores(<code>--total-executor-cores</code>) as expected resources to judge whether SchedulerBackend is ready.
Author: li-zhihui <zhihui.li@intel.com>
Author: Li Zhihui <zhihui.li@intel.com>
Closes#1525 from li-zhihui/fixre4s and squashes the following commits:
e9a630b [Li Zhihui] Rename variable totalExecutors and clean codes
abf4860 [Li Zhihui] Push down variable totalExpectedResources to children classes
ca54bd9 [li-zhihui] Format log with String interpolation
88c7dc6 [li-zhihui] Few codes and docs refactor
41cf47e [li-zhihui] Fix race condition at SchedulerBackend.isReady in standalone mode
References to dependencies(j) for actually obtaining RDD parents are less common than I originally estimated. It does clarify UnionRDD (also will clarify some of my other PRs)
Use of firstParent[T] is ubiquitous, but not as sure that benefits from being replaced with parent(0)[T].
Author: Erik Erlandson <eerlands@redhat.com>
Closes#1858 from erikerlandson/spark-2911-pr2 and squashes the following commits:
7ffea74 [Erik Erlandson] [SPARK-2911] apply parent[T](j) to clarify UnionRDD code
The decision order of Worker's WebUI port is "--webui-port", SPARK_WORKER_WEBUI_POR, 8081(default), spark.worker.ui.port. But in Master, the order is "--webui-port", spark.master.ui.port, SPARK_MASTER_WEBUI_PORT and 8080(default).
So we change the order in Worker's to keep it consistent with Master.
Author: WangTao <barneystinson@aliyun.com>
Closes#1838 from WangTaoTheTonic/reOrder and squashes the following commits:
460f4d4 [WangTao] Make decision order of Worker's WebUI consistent with Master's
Author: GuoQiang Li <witgo@qq.com>
Closes#1836 from witgo/SPARK-2897 and squashes the following commits:
23cdc5b [GuoQiang Li] review commit
ada4fba [GuoQiang Li] TorrentBroadcast does not support broadcast compression
fb91792 [GuoQiang Li] org.apache.spark.broadcast.TorrentBroadcast does use the serializer class specified in the spark option "spark.serializer"
Author: Erik Erlandson <eerlands@redhat.com>
Closes#1841 from erikerlandson/spark-2911-pr and squashes the following commits:
4699e2f [Erik Erlandson] [SPARK-2911]: provide rdd.parent[T](j) to obtain jth parent RDD
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#1834 from sarutak/SPARK-2904 and squashes the following commits:
38e7d45 [Kousuke Saruta] Removed non-used variable in SparkSubmitArguments
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1507 from sryza/sandy-spark-2565 and squashes the following commits:
74dad41 [Sandy Ryza] SPARK-2565. Update ShuffleReadMetrics as blocks are fetched
As described in https://issues.apache.org/jira/browse/SPARK-2787, right now sort-based shuffle is more expensive than hash-based for map operations that do no partial aggregation or sorting, such as groupByKey. This is because it has to serialize each data item twice (once when spilling to intermediate files, and then again when merging these files object-by-object). This patch adds a code path to just write separate files directly if the # of output partitions is small, and concatenate them at the end to produce a sorted file.
On the unit test side, I added some tests that force or don't force this bypass path to be used, and checked that our tests for other features (e.g. all the operations) cover both cases.
Author: Matei Zaharia <matei@databricks.com>
Closes#1799 from mateiz/SPARK-2787 and squashes the following commits:
88cf26a [Matei Zaharia] Fix rebase
10233af [Matei Zaharia] Review comments
398cb95 [Matei Zaharia] Fix looking up shuffle manager in conf
ca3efd9 [Matei Zaharia] Add docs for shuffle manager properties, and allow short names for them
d0ae3c5 [Matei Zaharia] Fix some comments
90d084f [Matei Zaharia] Add code path to bypass merge-sort in ExternalSorter, and tests
31e5d7c [Matei Zaharia] Move existing logic for writing partitioned files into ExternalSorter
fix bug of countApproxDistinct() when have more than one partition
Author: Davies Liu <davies.liu@gmail.com>
Closes#1812 from davies/approx and squashes the following commits:
bf757ce [Davies Liu] fix bug of countApproxDistinct() when have more than one partition
This patch modifies the ConnectionManager so that error messages are sent in reply when uncaught exceptions occur during message processing. This prevents message senders from hanging while waiting for an acknowledgment if the remote message processing failed.
This is an updated version of sarutak's PR, #1490. The main change is to use Futures / Promises to signal errors.
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Author: Josh Rosen <joshrosen@apache.org>
Closes#1758 from JoshRosen/connection-manager-fixes and squashes the following commits:
68620cb [Josh Rosen] Fix test in BlockFetcherIteratorSuite:
83673de [Josh Rosen] Error ACKs should trigger IOExceptions, so catch only those exceptions in the test.
b8bb4d4 [Josh Rosen] Fix manager.id vs managerServer.id typo that broke security tests.
659521f [Josh Rosen] Include previous exception when throwing new one
a2f745c [Josh Rosen] Remove sendMessageReliablySync; callers can wait themselves.
c01c450 [Josh Rosen] Return Try[Message] from sendMessageReliablySync.
f1cd1bb [Josh Rosen] Clean up @sarutak's PR #1490 for [SPARK-2583]: ConnectionManager error reporting
7399c6b [Josh Rosen] Merge remote-tracking branch 'origin/pr/1490' into connection-manager-fixes
ee91bb7 [Kousuke Saruta] Modified BufferMessage.scala to keep the spark code style
9dfd0d8 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
e7d9aa6 [Kousuke Saruta] rebase to master
326a17f [Kousuke Saruta] Add test cases to ConnectionManagerSuite.scala for SPARK-2583
2a18d6b [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
22d7ebd [Kousuke Saruta] Add test cases to BlockManagerSuite for SPARK-2583
e579302 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
281589c [Kousuke Saruta] Add a test case to BlockFetcherIteratorSuite.scala for fetching block from remote from successfully
0654128 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
ffaa83d [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
12d3de8 [Kousuke Saruta] Added BlockFetcherIteratorSuite.scala
4117b8f [Kousuke Saruta] Modified ConnectionManager to be alble to handle error during processing message
717c9c3 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
6635467 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2583
e2b8c4a [Kousuke Saruta] Modify to propagete error using ConnectionManager
I haven't tested this out on a cluster yet, but wanted to make sure the approach (passing ShuffleWriteMetrics down to DiskBlockObjectWriter) was ok
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1481 from sryza/sandy-spark-2566 and squashes the following commits:
8090d88 [Sandy Ryza] Fix ExternalSorter
b2a62ed [Sandy Ryza] Fix more test failures
8be6218 [Sandy Ryza] Fix test failures and mark a couple variables private
c5e68e5 [Sandy Ryza] SPARK-2566. Update ShuffleWriteMetrics incrementally
JIRA issues:
- Main: [SPARK-2678](https://issues.apache.org/jira/browse/SPARK-2678)
- Related: [SPARK-2874](https://issues.apache.org/jira/browse/SPARK-2874)
Related PR:
- #1715
This PR is both a fix for SPARK-2874 and a workaround for SPARK-2678. Fixing SPARK-2678 completely requires some API level changes that need further discussion, and we decided not to include it in Spark 1.1 release. As currently SPARK-2678 only affects Spark SQL scripts, this workaround is enough for Spark 1.1. Command line option handling logic in bash scripts looks somewhat dirty and duplicated, but it helps to provide a cleaner user interface as well as retain full downward compatibility for now.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes#1801 from liancheng/spark-2874 and squashes the following commits:
8045d7a [Cheng Lian] Make sure test suites pass
8493a9e [Cheng Lian] Using eval to retain quoted arguments
aed523f [Cheng Lian] Fixed typo in bin/spark-sql
f12a0b1 [Cheng Lian] Worked arount SPARK-2678
daee105 [Cheng Lian] Fixed usage messages of all Spark SQL related scripts
The goal of this PR is to allow users of Spark to write tight firewall rules for their clusters. This is currently not possible because Spark uses random ports in many places, notably the communication between executors and drivers. The changes in this PR are based on top of ash211's changes in #1107.
The list covered here may or may not be the complete set of port needed for Spark to operate perfectly. However, as of the latest commit there are no known sources of random ports (except in tests). I have not documented a few of the more obscure configs.
My spark-env.sh looks like this:
```
export SPARK_MASTER_PORT=6060
export SPARK_WORKER_PORT=7070
export SPARK_MASTER_WEBUI_PORT=9090
export SPARK_WORKER_WEBUI_PORT=9091
```
and my spark-defaults.conf looks like this:
```
spark.master spark://andrews-mbp:6060
spark.driver.port 5001
spark.fileserver.port 5011
spark.broadcast.port 5021
spark.replClassServer.port 5031
spark.blockManager.port 5041
spark.executor.port 5051
```
Author: Andrew Or <andrewor14@gmail.com>
Author: Andrew Ash <andrew@andrewash.com>
Closes#1777 from andrewor14/configure-ports and squashes the following commits:
621267b [Andrew Or] Merge branch 'master' of github.com:apache/spark into configure-ports
8a6b820 [Andrew Or] Use a random UI port during tests
7da0493 [Andrew Or] Fix tests
523c30e [Andrew Or] Add test for isBindCollision
b97b02a [Andrew Or] Minor fixes
c22ad00 [Andrew Or] Merge branch 'master' of github.com:apache/spark into configure-ports
93d359f [Andrew Or] Executors connect to wrong port when collision occurs
d502e5f [Andrew Or] Handle port collisions when creating Akka systems
a2dd05c [Andrew Or] Patrick's comment nit
86461e2 [Andrew Or] Remove spark.executor.env.port and spark.standalone.client.port
1d2d5c6 [Andrew Or] Fix ports for standalone cluster mode
cb3be88 [Andrew Or] Various doc fixes (broken link, format etc.)
e837cde [Andrew Or] Remove outdated TODOs
bfbab28 [Andrew Or] Merge branch 'master' of github.com:apache/spark into configure-ports
de1b207 [Andrew Or] Update docs to reflect new ports
b565079 [Andrew Or] Add spark.ports.maxRetries
2551eb2 [Andrew Or] Remove spark.worker.watcher.port
151327a [Andrew Or] Merge branch 'master' of github.com:apache/spark into configure-ports
9868358 [Andrew Or] Add a few miscellaneous ports
6016e77 [Andrew Or] Add spark.executor.port
8d836e6 [Andrew Or] Also document SPARK_{MASTER/WORKER}_WEBUI_PORT
4d9e6f3 [Andrew Or] Fix super subtle bug
3f8e51b [Andrew Or] Correct erroneous docs...
e111d08 [Andrew Or] Add names for UI services
470f38c [Andrew Or] Special case non-"Address already in use" exceptions
1d7e408 [Andrew Or] Treat 0 ports specially + return correct ConnectionManager port
ba32280 [Andrew Or] Minor fixes
6b550b0 [Andrew Or] Assorted fixes
73fbe89 [Andrew Or] Move start service logic to Utils
ec676f4 [Andrew Or] Merge branch 'SPARK-2157' of github.com:ash211/spark into configure-ports
038a579 [Andrew Ash] Trust the server start function to report the port the service started on
7c5bdc4 [Andrew Ash] Fix style issue
0347aef [Andrew Ash] Unify port fallback logic to a single place
24a4c32 [Andrew Ash] Remove type on val to match surrounding style
9e4ad96 [Andrew Ash] Reformat for style checker
5d84e0e [Andrew Ash] Document new port configuration options
066dc7a [Andrew Ash] Fix up HttpServer port increments
cad16da [Andrew Ash] Add fallover increment logic for HttpServer
c5a0568 [Andrew Ash] Fix ConnectionManager to retry with increment
b80d2fd [Andrew Ash] Make Spark's block manager port configurable
17c79bb [Andrew Ash] Add a configuration option for spark-shell's class server
f34115d [Andrew Ash] SPARK-1176 Add port configuration for HttpBroadcast
49ee29b [Andrew Ash] SPARK-1174 Add port configuration for HttpFileServer
1c0981a [Andrew Ash] Make port in HttpServer configurable
copied from original JIRA (https://issues.apache.org/jira/browse/SPARK-2294):
If an executor E is free, a task may be speculatively assigned to E when there are other tasks in the job that have not been launched (at all) yet. Similarly, a task without any locality preferences may be assigned to E when there was another NODE_LOCAL task that could have been scheduled.
This happens because TaskSchedulerImpl calls TaskSetManager.resourceOffer (which in turn calls TaskSetManager.findTask) with increasing locality levels, beginning with PROCESS_LOCAL, followed by NODE_LOCAL, and so on until the highest currently allowed level. Now, supposed NODE_LOCAL is the highest currently allowed locality level. The first time findTask is called, it will be called with max level PROCESS_LOCAL; if it cannot find any PROCESS_LOCAL tasks, it will try to schedule tasks with no locality preferences or speculative tasks. As a result, speculative tasks or tasks with no preferences may be scheduled instead of NODE_LOCAL tasks.
----
I added an additional parameter in resourceOffer and findTask, maxLocality, indicating when we should consider the tasks without locality preference
Author: CodingCat <zhunansjtu@gmail.com>
Closes#1313 from CodingCat/SPARK-2294 and squashes the following commits:
bf3f13b [CodingCat] rollback some forgotten changes
89f9bc0 [CodingCat] address matei's comments
18cae02 [CodingCat] add test case for node-local tasks
2ba6195 [CodingCat] fix failed test cases
87dd09e [CodingCat] fix style
9b9432f [CodingCat] remove hasNodeLocalOnlyTasks
fdd1573 [CodingCat] fix failed test cases
941a4fd [CodingCat] see my shocked face..........
f600085 [CodingCat] remove hasNodeLocalOnlyTasks checking
0b8a46b [CodingCat] test whether hasNodeLocalOnlyTasks affect the results
73ceda8 [CodingCat] style fix
b3a430b [CodingCat] remove fine granularity tracking for node-local only tasks
f9a2ad8 [CodingCat] simplify the logic in TaskSchedulerImpl
c8c1de4 [CodingCat] simplify the patch
be652ed [CodingCat] avoid unnecessary delay when we only have nopref tasks
dee9e22 [CodingCat] fix locality inversion bug in TaskManager by moving nopref branch
I inquired on dev mailing list about the motivation for checking the jdbc statement instead of the connection in the close() logic of JdbcRDD. Ted Yu believes there essentially is none- it is a simple cut and paste issue. So here is the tiny fix to patch it.
Author: Stephen Boesch <javadba>
Author: Stephen Boesch <javadba@gmail.com>
Closes#1792 from javadba/closejdbc and squashes the following commits:
363be4f [Stephen Boesch] SPARK-2869 - Fix tiny bug in JdbcRdd for closing jdbc connection (reformat with braces)
6518d36 [Stephen Boesch] SPARK-2689 Fix tiny bug in JdbcRdd for closing jdbc connection
3fb23ed [Stephen Boesch] SPARK-2689 Fix potential leak of connection/PreparedStatement in case of error in JdbcRDD
095b2c9 [Stephen Boesch] Fix tiny bug (likely copy and paste error) in closing jdbc connection
This can substantially reduce memory usage during shuffle.
Author: Reynold Xin <rxin@apache.org>
Closes#1781 from rxin/SPARK-2503-spark.shuffle.file.buffer.kb and squashes the following commits:
104b8d8 [Reynold Xin] [SPARK-2503] Lower shuffle output buffer (spark.shuffle.file.buffer.kb) to 32KB.
This patch adds support for giving accumulators user-visible names and displaying accumulator values in the web UI. This allows users to create custom counters that can display in the UI. The current approach displays both the accumulator deltas caused by each task and a "current" value of the accumulator totals for each stage, which gets update as tasks finish.
Currently in Spark developers have been extending the `TaskMetrics` functionality to provide custom instrumentation for RDD's. This provides a potentially nicer alternative of going through the existing accumulator framework (actually `TaskMetrics` and accumulators are on an awkward collision course as we add more features to the former). The current patch demo's how we can use the feature to provide instrumentation for RDD input sizes. The nice thing about going through accumulators is that users can read the current value of the data being tracked in their programs. This could be useful to e.g. decide to short-circuit a Spark stage depending on how things are going.
Author: Patrick Wendell <pwendell@gmail.com>
Closes#1309 from pwendell/metrics and squashes the following commits:
8815308 [Patrick Wendell] Merge remote-tracking branch 'apache/master' into HEAD
93fbe0f [Patrick Wendell] Other minor fixes
cc43f68 [Patrick Wendell] Updating unit tests
c991b1b [Patrick Wendell] Moving some code into the Accumulators class
9a9ba3c [Patrick Wendell] More merge fixes
c5ace9e [Patrick Wendell] More merge conflicts
1da15e3 [Patrick Wendell] Merge remote-tracking branch 'apache/master' into metrics
9860c55 [Patrick Wendell] Potential solution to posting listener events
0bb0e33 [Patrick Wendell] Remove "display" variable and assume display = name.isDefined
0ec4ac7 [Patrick Wendell] Java API's
e95bf69 [Patrick Wendell] Stash
be97261 [Patrick Wendell] Style fix
8407308 [Patrick Wendell] Removing examples in Hadoop and RDD class
64d405f [Patrick Wendell] Adding missing file
5d8b156 [Patrick Wendell] Changes based on Kay's review.
9f18bad [Patrick Wendell] Minor style changes and tests
7a63abc [Patrick Wendell] Adding Json serialization and responding to Reynold's feedback
ad85076 [Patrick Wendell] Example of using named accumulators for custom RDD metrics.
0b72660 [Patrick Wendell] Initial WIP example of supporing globally named accumulators.
It was easier to combine these 2 jira since they touch many of the same places. This pr adds the following:
- adds modify acls
- adds admin acls (list of admins/users that get added to both view and modify acls)
- modify Kill button on UI to take modify acls into account
- changes config name of spark.ui.acls.enable to spark.acls.enable since I choose poorly in original name. We keep backwards compatibility so people can still use spark.ui.acls.enable. The acls should apply to any web ui as well as any CLI interfaces.
- send view and modify acls information on to YARN so that YARN interfaces can use (yarn cli for killing applications for example).
Author: Thomas Graves <tgraves@apache.org>
Closes#1196 from tgravescs/SPARK-1890 and squashes the following commits:
8292eb1 [Thomas Graves] review comments
b92ec89 [Thomas Graves] remove unneeded variable from applistener
4c765f4 [Thomas Graves] Add in admin acls
72eb0ac [Thomas Graves] Add modify acls
Author: Reynold Xin <rxin@apache.org>
Closes#1780 from rxin/kryo-init-size and squashes the following commits:
551b935 [Reynold Xin] [SPARK-2856] Decrease initial buffer size for Kryo to 64KB.
`master.ui.port` and `worker.ui.port` were never picked up by SparkConf, simply because they are not prefixed with "spark." Unfortunately, this is also currently the documented way of setting these values.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1779 from andrewor14/master-worker-port and squashes the following commits:
8475e95 [Andrew Or] Update docs to reflect changes in configs
4db3d5d [Andrew Or] Stop using configs that don't actually work
This tracks memory properly if there are multiple spilling collections in the same task (which was a problem before), and also implements an algorithm that lets each thread grow up to 1 / 2N of the memory pool (where N is the number of threads) before spilling, which avoids an inefficiency with small spills we had before (some threads would spill many times at 0-1 MB because the pool was allocated elsewhere).
Author: Matei Zaharia <matei@databricks.com>
Closes#1707 from mateiz/spark-2711 and squashes the following commits:
debf75b [Matei Zaharia] Review comments
24f28f3 [Matei Zaharia] Small rename
c8f3a8b [Matei Zaharia] Update ShuffleMemoryManager to be able to partially grant requests
315e3a5 [Matei Zaharia] Some review comments
b810120 [Matei Zaharia] Create central manager to track memory for all spilling collections
Replaces this with an O(1) operation that does not have to shift over
the whole tail of the array into the gap produced by the element removed.
Author: Matei Zaharia <matei@databricks.com>
Closes#1773 from mateiz/SPARK-2685 and squashes the following commits:
1ea028a [Matei Zaharia] Update comments in StreamBuffer and EAOM, and reuse ArrayBuffers
eb1abfd [Matei Zaharia] Update ExternalAppendOnlyMap to avoid buffer.remove()
Author: Reynold Xin <rxin@apache.org>
Closes#1772 from rxin/accumulator-dagscheduler and squashes the following commits:
6a58520 [Reynold Xin] [SPARK-2323] Exception in accumulator update should not crash DAGScheduler & SparkContext.
All these changes are from mridulm's work in #1609, but extracted here to fix this specific issue and make it easier to merge not 1.1. This particular set of changes is to make sure that we read exactly the right range of bytes from each spill file in EAOM: some serializers can write bytes after the last object (e.g. the TC_RESET flag in Java serialization) and that would confuse the previous code into reading it as part of the next batch. There are also improvements to cleanup to make sure files are closed.
In addition to bringing in the changes to ExternalAppendOnlyMap, I also copied them to the corresponding code in ExternalSorter and updated its test suite to test for the same issues.
Author: Matei Zaharia <matei@databricks.com>
Closes#1722 from mateiz/spark-2792 and squashes the following commits:
5d4bfb5 [Matei Zaharia] Make objectStreamReset counter count the last object written too
18fe865 [Matei Zaharia] Update docs on objectStreamReset
576ee83 [Matei Zaharia] Allow objectStreamReset to be 0
0374217 [Matei Zaharia] Remove super paranoid code to close file handles
bda37bb [Matei Zaharia] Implement Mridul's ExternalAppendOnlyMap fixes in ExternalSorter too
0d6dad7 [Matei Zaharia] Added Mridul's test changes for ExternalAppendOnlyMap
9a78e4b [Matei Zaharia] Add @mridulm's fixes to ExternalAppendOnlyMap for batch sizes
Kill only the python worker related to cancelled tasks.
The daemon will start a background thread to monitor all the opened sockets for all workers. If the socket is closed by JVM, this thread will kill the worker.
When an task is cancelled, the socket to worker will be closed, then the worker will be killed by deamon.
Author: Davies Liu <davies.liu@gmail.com>
Closes#1643 from davies/kill and squashes the following commits:
8ffe9f3 [Davies Liu] kill worker by deamon, because runtime.exec() is too heavy
46ca150 [Davies Liu] address comment
acd751c [Davies Liu] kill the worker when task is canceled
Minor fixes on top of #1679.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1736 from andrewor14/amend-#1679 and squashes the following commits:
3b46f5e [Andrew Or] Minor fixes
This also Closes#1701.
Author: GuoQiang Li <witgo@qq.com>
Closes#1208 from witgo/SPARK-1470 and squashes the following commits:
422646b [GuoQiang Li] Remove scalalogging-slf4j dependency
`spark.test.home` was introduced in #1734. This is fine for SBT but is failing maven tests. Either way it shouldn't throw an NPE.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1739 from andrewor14/fix-spark-test-home and squashes the following commits:
ce2624c [Andrew Or] Do not throw NPE if spark.test.home is not set
When standalone Workers launch executors, they inherit the Spark home set by the driver. This means if the worker machines do not share the same directory structure as the driver node, the Workers will attempt to run scripts (e.g. bin/compute-classpath.sh) that do not exist locally and fail. This is a common scenario if the driver is launched from outside of the cluster.
The solution is to simply not pass the driver's Spark home to the Workers. This PR further makes an attempt to avoid overloading the usages of `spark.home`, which is now only used for setting executor Spark home on Mesos and in python.
This is based on top of #1392 and originally reported by YanTangZhai. Tested on standalone cluster.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1734 from andrewor14/spark-home-reprise and squashes the following commits:
f71f391 [Andrew Or] Revert changes in python
1c2532c [Andrew Or] Merge branch 'master' of github.com:apache/spark into spark-home-reprise
188fc5d [Andrew Or] Avoid using spark.home where possible
09272b7 [Andrew Or] Always use Worker's working directory as spark home
The existing code in `StorageUtils` is not the most efficient. Every time we want to update an `RDDInfo` we end up iterating through all blocks on all block managers just to discard most of them. The symptoms manifest themselves in the bountiful UI bugs observed in the wild. Many of these bugs are caused by the slow consumption of events in `LiveListenerBus`, which frequently leads to the event queue overflowing and `SparkListenerEvent`s being dropped on the floor. The changes made in this PR avoid this by first filtering out only the blocks relevant to us before computing storage information from them.
It's worth a mention that this corner of the Spark code is also not very well-tested at all. The bulk of the changes in this PR (more than 60%) is actually test cases for the various logic in `StorageUtils.scala` as well as `StorageTab.scala`. These will eventually be extended to cover the various listeners that constitute the `SparkUI`.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1679 from andrewor14/fix-drop-events and squashes the following commits:
f80c1fa [Andrew Or] Rewrite fold and reduceOption as sum
e132d69 [Andrew Or] Merge branch 'master' of github.com:apache/spark into fix-drop-events
14fa1c3 [Andrew Or] Simplify some code + update a few comments
a91be46 [Andrew Or] Make ExecutorsPage blazingly fast
bf6f09b [Andrew Or] Minor changes
8981de1 [Andrew Or] Merge branch 'master' of github.com:apache/spark into fix-drop-events
af19bc0 [Andrew Or] *UsedByRDD -> *UsedByRdd (minor)
6970bc8 [Andrew Or] Add extensive tests for StorageListener and the new code in StorageUtils
e080b9e [Andrew Or] Reduce run time of StorageUtils.updateRddInfo to near constant
2c3ef6a [Andrew Or] Actually filter out only the relevant RDDs
6fef86a [Andrew Or] Add extensive tests for new code in StorageStatus
b66b6b0 [Andrew Or] Use more efficient underlying data structures for blocks
6a7b7c0 [Andrew Or] Avoid chained operations on TraversableLike
a9ec384 [Andrew Or] Merge branch 'master' of github.com:apache/spark into fix-drop-events
b12fcd7 [Andrew Or] Fix tests + simplify sc.getRDDStorageInfo
da8e322 [Andrew Or] Merge branch 'master' of github.com:apache/spark into fix-drop-events
8e91921 [Andrew Or] Iterate through a filtered set of blocks when updating RDDInfo
7b2c4aa [Andrew Or] Rewrite blockLocationsFromStorageStatus + clean up method signatures
41fa50d [Andrew Or] Add a legacy constructor for StorageStatus
53af15d [Andrew Or] Refactor StorageStatus + add a bunch of tests
Author: GuoQiang Li <witgo@qq.com>
Closes#1369 from witgo/SPARK-1470_new and squashes the following commits:
66a1641 [GuoQiang Li] IncompatibleResultTypeProblem
73a89ba [GuoQiang Li] Use the scala-logging wrapper instead of the directly sfl4j api.
Curently, daemon.py forks a pool of numProcessors subprocesses, and those processes fork themselves again to create the actual Python worker processes that handle data.
I think that this extra layer of indirection is unnecessary and adds a lot of complexity. This commit attempts to remove this middle layer of subprocesses by launching the workers directly from daemon.py.
See https://github.com/mesos/spark/pull/563 for the original PR that added daemon.py, where I raise some issues with the current design.
Author: Josh Rosen <joshrosen@apache.org>
Closes#1680 from JoshRosen/pyspark-daemon and squashes the following commits:
5abbcb9 [Josh Rosen] Replace magic number: 4 -> EINTR
5495dff [Josh Rosen] Throw IllegalStateException if worker launch fails.
b79254d [Josh Rosen] Detect failed fork() calls; improve error logging.
282c2c4 [Josh Rosen] Remove daemon.py exit logging, since it caused problems:
8554536 [Josh Rosen] Fix daemon’s shutdown(); log shutdown reason.
4e0fab8 [Josh Rosen] Remove shared-memory exit_flag; don't die on worker death.
e9892b4 [Josh Rosen] [WIP] [SPARK-2764] Simplify daemon.py process structure.
If SPARK_CONF_DIR environment variable is set, search it for spark-defaults.conf.
Author: Albert Chu <chu11@llnl.gov>
Closes#1059 from chu11/SPARK-2116 and squashes the following commits:
9f3ac94 [Albert Chu] SPARK-2116: If SPARK_CONF_DIR environment variable is set, search it for spark-defaults.conf.
Convert Row in JavaSchemaRDD into Array[Any] and unpickle them as tuple in Python, then convert them into namedtuple, so use can access fields just like attributes.
This will let nested structure can be accessed as object, also it will reduce the size of serialized data and better performance.
root
|-- field1: integer (nullable = true)
|-- field2: string (nullable = true)
|-- field3: struct (nullable = true)
| |-- field4: integer (nullable = true)
| |-- field5: array (nullable = true)
| | |-- element: integer (containsNull = false)
|-- field6: array (nullable = true)
| |-- element: struct (containsNull = false)
| | |-- field7: string (nullable = true)
Then we can access them by row.field3.field5[0] or row.field6[5].field7
It also will infer the schema in Python, convert Row/dict/namedtuple/objects into tuple before serialization, then call applySchema in JVM. During inferSchema(), the top level of dict in row will be StructType, but any nested dictionary will be MapType.
You can use pyspark.sql.Row to convert unnamed structure into Row object, make the RDD can be inferable. Such as:
ctx.inferSchema(rdd.map(lambda x: Row(a=x[0], b=x[1]))
Or you could use Row to create a class just like namedtuple, for example:
Person = Row("name", "age")
ctx.inferSchema(rdd.map(lambda x: Person(*x)))
Also, you can call applySchema to apply an schema to a RDD of tuple/list and turn it into a SchemaRDD. The `schema` should be StructType, see the API docs for details.
schema = StructType([StructField("name, StringType, True),
StructType("age", IntegerType, True)])
ctx.applySchema(rdd, schema)
PS: In order to use namedtuple to inferSchema, you should make namedtuple picklable.
Author: Davies Liu <davies.liu@gmail.com>
Closes#1598 from davies/nested and squashes the following commits:
f1d15b6 [Davies Liu] verify schema with the first few rows
8852aaf [Davies Liu] check type of schema
abe9e6e [Davies Liu] address comments
61b2292 [Davies Liu] add @deprecated to pythonToJavaMap
1e5b801 [Davies Liu] improve cache of classes
51aa135 [Davies Liu] use Row to infer schema
e9c0d5c [Davies Liu] remove string typed schema
353a3f2 [Davies Liu] fix code style
63de8f8 [Davies Liu] fix typo
c79ca67 [Davies Liu] fix serialization of nested data
6b258b5 [Davies Liu] fix pep8
9d8447c [Davies Liu] apply schema provided by string of names
f5df97f [Davies Liu] refactor, address comments
9d9af55 [Davies Liu] use arrry to applySchema and infer schema in Python
84679b3 [Davies Liu] Merge branch 'master' of github.com:apache/spark into nested
0eaaf56 [Davies Liu] fix doc tests
b3559b4 [Davies Liu] use generated Row instead of namedtuple
c4ddc30 [Davies Liu] fix conflict between name of fields and variables
7f6f251 [Davies Liu] address all comments
d69d397 [Davies Liu] refactor
2cc2d45 [Davies Liu] refactor
182fb46 [Davies Liu] refactor
bc6e9e1 [Davies Liu] switch to new Schema API
547bf3e [Davies Liu] Merge branch 'master' into nested
a435b5a [Davies Liu] add docs and code refactor
2c8debc [Davies Liu] Merge branch 'master' into nested
644665a [Davies Liu] use tuple and namedtuple for schemardd
All changes from this PR are by mridulm and are drawn from his work in #1609. This patch is intended to fix all major issues related to shuffle file consolidation that mridulm found, while minimizing changes to the code, with the hope that it may be more easily merged into 1.1.
This patch is **not** intended as a replacement for #1609, which provides many additional benefits, including fixes to ExternalAppendOnlyMap, improvements to DiskBlockObjectWriter's API, and several new unit tests.
If it is feasible to merge #1609 for the 1.1 deadline, that is a preferable option.
Author: Aaron Davidson <aaron@databricks.com>
Closes#1678 from aarondav/consol and squashes the following commits:
53b3f6d [Aaron Davidson] Correct behavior when writing unopened file
701d045 [Aaron Davidson] Rebase with sort-based shuffle
9160149 [Aaron Davidson] SPARK-2532: Minimal shuffle consolidation fixes
JIRA: https://issues.apache.org/jira/browse/SPARK-1612
Move the "close" statements into a "finally" block.
Author: zsxwing <zsxwing@gmail.com>
Closes#535 from zsxwing/SPARK-1612 and squashes the following commits:
ae52f50 [zsxwing] Update to follow the code style
549ba13 [zsxwing] SPARK-1612: Fix potential resource leaks
When performing some transformations on RDDs after many iterations, the dependencies of RDDs could be very long. It can easily cause StackOverflowError when recursively visiting these dependencies in Spark core. For example:
var rdd = sc.makeRDD(Array(1))
for (i <- 1 to 1000) {
rdd = rdd.coalesce(1).cache()
rdd.collect()
}
This PR changes recursive visiting on rdd's dependencies to iterative approach to avoid StackOverflowError.
In addition to the recursive visiting, since the Java serializer has a known [bug](http://bugs.java.com/bugdatabase/view_bug.do?bug_id=4152790) that causes StackOverflowError too when serializing/deserializing a large graph of objects. So applying this PR only solves part of the problem. Using KryoSerializer to replace Java serializer might be helpful. However, since KryoSerializer is not supported for `spark.closure.serializer` now, I can not test if KryoSerializer can solve Java serializer's problem completely.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#1418 from viirya/remove_recursive_visit and squashes the following commits:
6b2c615 [Liang-Chi Hsieh] change function name; comply with code style.
5f072a7 [Liang-Chi Hsieh] add comments to explain Stack usage.
8742dbb [Liang-Chi Hsieh] comply with code style.
900538b [Liang-Chi Hsieh] change recursive visiting on rdd's dependencies to iterative approach to avoid stackoverflowerror.
getPreferredLocs traverses a dependency graph of partitions using depth first search. Given a complex dependency graph, the old implementation may explore a set of paths in the graph that is exponential in the number of nodes. By maintaining a set of visited nodes the new implementation avoids revisiting nodes, preventing exponential blowup.
Some comment and whitespace cleanups are also included.
Author: Aaron Staple <aaron.staple@gmail.com>
Closes#1362 from staple/SPARK-695 and squashes the following commits:
ecea0f3 [Aaron Staple] address review comments
751c661 [Aaron Staple] [SPARK-695] Add a unit test.
5adf326 [Aaron Staple] Replace getPreferredLocsInternal's HashMap argument with a simpler HashSet.
58e37d0 [Aaron Staple] Replace comment documenting NarrowDependency.
6751ced [Aaron Staple] Revert "Remove unused variable."
04c7097 [Aaron Staple] Fix indentation.
0030884 [Aaron Staple] Remove unused variable.
33f67c6 [Aaron Staple] Clarify comment.
4e42b46 [Aaron Staple] Remove apparently incorrect comment describing NarrowDependency.
65c2d3d [Aaron Staple] [SPARK-695] In DAGScheduler's getPreferredLocs, track set of visited partitions.
This is a sketch of a patch that allows the UI to show metrics for tasks that have not yet completed. It adds a heartbeat every 2 seconds from the executors to the driver, reporting metrics for all of the executor's tasks.
It still needs unit tests, polish, and cluster testing, but I wanted to put it up to get feedback on the approach.
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1056 from sryza/sandy-spark-2099 and squashes the following commits:
93b9fdb [Sandy Ryza] Up heartbeat interval to 10 seconds and other tidying
132aec7 [Sandy Ryza] Heartbeat and HeartbeatResponse are already Serializable as case classes
38dffde [Sandy Ryza] Additional review feedback and restore test that was removed in BlockManagerSuite
51fa396 [Sandy Ryza] Remove hostname race, add better comments about threading, and some stylistic improvements
3084f10 [Sandy Ryza] Make TaskUIData a case class again
3bda974 [Sandy Ryza] Stylistic fixes
0dae734 [Sandy Ryza] SPARK-2099. Report progress while task is running.
[SPARK-2557](https://issues.apache.org/jira/browse/SPARK-2557)
Author: Ye Xianjin <advancedxy@gmail.com>
Closes#1464 from advancedxy/SPARK-2557 and squashes the following commits:
d844d67 [Ye Xianjin] add local-*-n-failures, bad-local-n, bad-local-n-failures test case
3bbc668 [Ye Xianjin] fix LOCAL_N_REGEX regular expression and make local_n_failures accept * as all cores on the computer
Author: Rahul Singhal <rahul.singhal@guavus.com>
Closes#1076 from rahulsinghaliitd/SPARK-2134 and squashes the following commits:
15f18b6 [Rahul Singhal] SPARK-2134: Report metrics before application finishes
This patch simply uses the ExternalSorter class from sort-based shuffle.
Closes#931 and Closes#1090
Author: Matei Zaharia <matei@databricks.com>
Closes#1677 from mateiz/spark-983 and squashes the following commits:
96b3fda [Matei Zaharia] SPARK-983. Support external sorting in sortByKey()
Author: Kousuke Saruta <sarutak@oss.nttdata.co.jp>
Closes#1578 from sarutak/SPARK-2670 and squashes the following commits:
85c8938 [Kousuke Saruta] Removed useless results.put for fail fast
e8713cc [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2670
d353984 [Kousuke Saruta] Refined assertion messages in BlockFetcherIteratorSuite.scala
03bcb02 [Kousuke Saruta] Merge branch 'SPARK-2670' of github.com:sarutak/spark into SPARK-2670
5d05855 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2670
4fca130 [Kousuke Saruta] Added test cases for BasicBlockFetcherIterator
b7b8250 [Kousuke Saruta] Modified BasicBlockFetchIterator to fail fast when local fetch error has been occurred
a3a9be1 [Kousuke Saruta] Modified BlockFetcherIterator for SPARK-2670
460dc01 [Kousuke Saruta] Merge branch 'master' of git://git.apache.org/spark into SPARK-2670
e310c0b [Kousuke Saruta] Modified BlockFetcherIterator to handle local fetch failure as fatch fail
Prior to this change, every PySpark task completion opened a new socket to the accumulator server, passed its updates through, and then quit. I'm not entirely sure why PySpark always sends accumulator updates, but regardless this causes a very rapid buildup of ephemeral TCP connections that remain in the TCP_WAIT state for around a minute before being cleaned up.
Rather than trying to allow these sockets to be cleaned up faster, this patch simply reuses the connection between tasks completions (since they're fed updates in a single-threaded manner by the DAGScheduler anyway).
The only tricky part here was making sure that the AccumulatorServer was able to shutdown in a timely manner (i.e., stop polling for new data), and this was accomplished via minor feats of magic.
I have confirmed that this patch eliminates the buildup of ephemeral sockets due to the accumulator updates. However, I did note that there were still significant sockets being created against the PySpark daemon port, but my machine was not able to create enough sockets fast enough to fail. This may not be the last time we've seen this issue, though.
Author: Aaron Davidson <aaron@databricks.com>
Closes#1503 from aarondav/accum and squashes the following commits:
b3e12f7 [Aaron Davidson] SPARK-2282: Reuse Socket for sending accumulator updates to Pyspark
It should be more convenient if user can specify ascending and numPartitions when calling sortByKey.
Author: Rui Li <rui.li@intel.com>
Closes#1645 from lirui-intel/spark-2740 and squashes the following commits:
fb5d52e [Rui Li] SPARK-2740: allow user to specify ascending and numPartitions for sortByKey
The logging code that handles log4j initialization leads to an stack overflow error when used with log4j 2.x, which has just been released. This occurs even a downstream project has correctly adjusted SLF4J bindings, and that is the right thing to do for log4j 2.x, since it is effectively a separate project from 1.x.
Here is the relevant bit of Logging.scala:
```
private def initializeLogging() {
// If Log4j is being used, but is not initialized, load a default properties file
val binder = StaticLoggerBinder.getSingleton
val usingLog4j = binder.getLoggerFactoryClassStr.endsWith("Log4jLoggerFactory")
val log4jInitialized = LogManager.getRootLogger.getAllAppenders.hasMoreElements
if (!log4jInitialized && usingLog4j) {
val defaultLogProps = "org/apache/spark/log4j-defaults.properties"
Option(Utils.getSparkClassLoader.getResource(defaultLogProps)) match {
case Some(url) =>
PropertyConfigurator.configure(url)
log.info(s"Using Spark's default log4j profile: $defaultLogProps")
case None =>
System.err.println(s"Spark was unable to load $defaultLogProps")
}
}
Logging.initialized = true
// Force a call into slf4j to initialize it. Avoids this happening from mutliple threads
// and triggering this: http://mailman.qos.ch/pipermail/slf4j-dev/2010-April/002956.html
log
}
```
The first minor issue is that there is a call to a logger inside this method, which is initializing logging. In this situation, it ends up causing the initialization to be called recursively until the stack overflow. It would be slightly tidier to log this only after Logging.initialized = true. Or not at all. But it's not the root problem, or else, it would not work at all now.
The calls to log4j classes here always reference log4j 1.2 no matter what. For example, there is not getAllAppenders in log4j 2.x. That's fine. Really, "usingLog4j" means "using log4j 1.2" and "log4jInitialized" means "log4j 1.2 is initialized".
usingLog4j should be false for log4j 2.x, because the initialization only matters for log4j 1.2. But, it's true, and that's the real issue. And log4jInitialized is always false, since calls to the log4j 1.2 API are stubs and no-ops in this setup, where the caller has swapped in log4j 2.x. Hence the loop.
This is fixed, I believe, if "usingLog4j" can be false for log4j 2.x. The SLF4J static binding class has the same name for both versions, unfortunately, which causes the issue. However they're in different packages. For example, if the test included "... and begins with org.slf4j", it should work, as the SLF4J binding for log4j 2.x is provided by log4j 2.x at the moment, and is in package org.apache.logging.slf4j.
Of course, I assume that SLF4J will eventually offer its own binding. I hope to goodness they at least name the binding class differently, or else this will again not work. But then some other check can probably be made.
Author: Sean Owen <srowen@gmail.com>
Closes#1547 from srowen/SPARK-2646 and squashes the following commits:
92a9898 [Sean Owen] System.out -> System.err
94be4c7 [Sean Owen] Add back log message as System.out, with informational comment
a7f8876 [Sean Owen] Updates from review
6f3c1d3 [Sean Owen] Remove log statement in logging initialization, and distinguish log4j 1.2 from 2.0, to avoid stack overflow in initialization
...ags
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1665 from sryza/sandy-spark-2664 and squashes the following commits:
0518c63 [Sandy Ryza] SPARK-2664. Deal with `--conf` options in spark-submit that relate to flags
This allows users to gain access to the InputSplit which backs each partition.
An alternative solution would have been to have a .withInputSplit() method which returns a new RDD[(InputSplit, (K, V))], but this is confusing because you could not cache this RDD or shuffle it, as InputSplit is not inherently serializable.
Author: Aaron Davidson <aaron@databricks.com>
Closes#973 from aarondav/hadoop and squashes the following commits:
9c9112b [Aaron Davidson] Add JavaAPISuite test
9942cd7 [Aaron Davidson] Add Java API
1284a3a [Aaron Davidson] SPARK-2028: Expose mapPartitionsWithInputSplit in HadoopRDD
The Java API's use of fake ClassTags doesn't seem to cause any problems for Java users, but it can lead to issues when passing JavaRDDs' underlying RDDs to Scala code (e.g. in the MLlib Java API wrapper code). If we call collect() on a Scala RDD with an incorrect ClassTag, this causes ClassCastExceptions when we try to allocate an array of the wrong type (for example, see SPARK-2197).
There are a few possible fixes here. An API-breaking fix would be to completely remove the fake ClassTags and require Java API users to pass java.lang.Class instances to all parallelize() calls and add returnClass fields to all Function implementations. This would be extremely verbose.
Instead, this patch adds internal APIs to "repair" a Scala RDD with an incorrect ClassTag by wrapping it and overriding its ClassTag. This should be okay for cases where the Scala code that calls collect() knows what type of array should be allocated, which is the case in the MLlib wrappers.
Author: Josh Rosen <joshrosen@apache.org>
Closes#1639 from JoshRosen/SPARK-2737 and squashes the following commits:
572b4c8 [Josh Rosen] Replace newRDD[T] with mapPartitions().
469d941 [Josh Rosen] Preserve partitioner in retag().
af78816 [Josh Rosen] Allow retag() to get classTag implicitly.
d1d54e6 [Josh Rosen] [SPARK-2737] Add retag() method for changing RDDs' ClassTags.
We resolve relative paths to the local `file:/` system for `--jars` and `--files` in spark submit (#853). We should do the same for the history server.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1280 from andrewor14/hist-serv-fix and squashes the following commits:
13ff406 [Andrew Or] Merge branch 'master' of github.com:apache/spark into hist-serv-fix
b393e17 [Andrew Or] Strip trailing "/" from logging directory
622a471 [Andrew Or] Fix test in EventLoggingListenerSuite
0e20f71 [Andrew Or] Shift responsibility of resolving paths up one level
b037c0c [Andrew Or] Use resolved paths for everything in history server
c7e36ee [Andrew Or] Resolve paths for event logging too
40e3933 [Andrew Or] Resolve history server file paths
Author: Reynold Xin <rxin@apache.org>
Closes#1675 from rxin/unionrdd and squashes the following commits:
941d316 [Reynold Xin] Clear RDDs for checkpointing.
c9f05f2 [Reynold Xin] [SPARK-2758] UnionRDD's UnionPartition should not reference parent RDDs
This adds a new ShuffleManager based on sorting, as described in https://issues.apache.org/jira/browse/SPARK-2045. The bulk of the code is in an ExternalSorter class that is similar to ExternalAppendOnlyMap, but sorts key-value pairs by partition ID and can be used to create a single sorted file with a map task's output. (Longer-term I think this can take on the remaining functionality in ExternalAppendOnlyMap and replace it so we don't have code duplication.)
The main TODOs still left are:
- [x] enabling ExternalSorter to merge across spilled files
- [x] with an Ordering
- [x] without an Ordering, using the keys' hash codes
- [x] adding more tests (e.g. a version of our shuffle suite that runs on this)
- [x] rebasing on top of the size-tracking refactoring in #1165 when that is merged
- [x] disabling spilling if spark.shuffle.spill is set to false
Despite this though, this seems to work pretty well (running successfully in cases where the hash shuffle would OOM, such as 1000 reduce tasks on executors with only 1G memory), and it seems to be comparable in speed or faster than hash-based shuffle (it will create much fewer files for the OS to keep track of). So I'm posting it to get some early feedback.
After these TODOs are done, I'd also like to enable ExternalSorter to sort data within each partition by a key as well, which will allow us to use it to implement external spilling in reduce tasks in `sortByKey`.
Author: Matei Zaharia <matei@databricks.com>
Closes#1499 from mateiz/sort-based-shuffle and squashes the following commits:
bd841f9 [Matei Zaharia] Various review comments
d1c137fd [Matei Zaharia] Various review comments
a611159 [Matei Zaharia] Compile fixes due to rebase
62c56c8 [Matei Zaharia] Fix ShuffledRDD sometimes not returning Tuple2s.
f617432 [Matei Zaharia] Fix a failing test (seems to be due to change in SizeTracker logic)
9464d5f [Matei Zaharia] Simplify code and fix conflicts after latest rebase
0174149 [Matei Zaharia] Add cleanup behavior and cleanup tests for sort-based shuffle
eb4ee0d [Matei Zaharia] Remove customizable element type in ShuffledRDD
fa2e8db [Matei Zaharia] Allow nextBatchStream to be called after we're done looking at all streams
a34b352 [Matei Zaharia] Fix tracking of indices within a partition in SpillReader, and add test
03e1006 [Matei Zaharia] Add a SortShuffleSuite that runs ShuffleSuite with sort-based shuffle
3c7ff1f [Matei Zaharia] Obey the spark.shuffle.spill setting in ExternalSorter
ad65fbd [Matei Zaharia] Rebase on top of Aaron's Sorter change, and use Sorter in our buffer
44d2a93 [Matei Zaharia] Use estimateSize instead of atGrowThreshold to test collection sizes
5686f71 [Matei Zaharia] Optimize merging phase for in-memory only data:
5461cbb [Matei Zaharia] Review comments and more tests (e.g. tests with 1 element per partition)
e9ad356 [Matei Zaharia] Update ContextCleanerSuite to make sure shuffle cleanup tests use hash shuffle (since they were written for it)
c72362a [Matei Zaharia] Added bug fix and test for when iterators are empty
de1fb40 [Matei Zaharia] Make trait SizeTrackingCollection private[spark]
4988d16 [Matei Zaharia] tweak
c1b7572 [Matei Zaharia] Small optimization
ba7db7f [Matei Zaharia] Handle null keys in hash-based comparator, and add tests for collisions
ef4e397 [Matei Zaharia] Support for partial aggregation even without an Ordering
4b7a5ce [Matei Zaharia] More tests, and ability to sort data if a total ordering is given
e1f84be [Matei Zaharia] Fix disk block manager test
5a40a1c [Matei Zaharia] More tests
614f1b4 [Matei Zaharia] Add spill metrics to map tasks
cc52caf [Matei Zaharia] Add more error handling and tests for error cases
bbf359d [Matei Zaharia] More work
3a56341 [Matei Zaharia] More partial work towards sort-based shuffle
7a0895d [Matei Zaharia] Some more partial work towards sort-based shuffle
b615476 [Matei Zaharia] Scaffolding for sort-based shuffle
JIRA issue: https://issues.apache.org/jira/browse/SPARK-2024
This PR is a followup to #455 and adds capabilities for saving PySpark RDDs using SequenceFile or any Hadoop OutputFormats.
* Added RDD methods ```saveAsSequenceFile```, ```saveAsHadoopFile``` and ```saveAsHadoopDataset```, for both old and new MapReduce APIs.
* Default converter for converting common data types to Writables. Users may specify custom converters to convert to desired data types.
* No out-of-box support for reading/writing arrays, since ArrayWritable itself doesn't have a no-arg constructor for creating an empty instance upon reading. Users need to provide ArrayWritable subtypes. Custom converters for converting arrays to suitable ArrayWritable subtypes are also needed when writing. When reading, the default converter will convert any custom ArrayWritable subtypes to ```Object[]``` and they get pickled to Python tuples.
* Added HBase and Cassandra output examples to show how custom output formats and converters can be used.
cc MLnick mateiz ahirreddy pwendell
Author: Kan Zhang <kzhang@apache.org>
Closes#1338 from kanzhang/SPARK-2024 and squashes the following commits:
c01e3ef [Kan Zhang] [SPARK-2024] code formatting
6591e37 [Kan Zhang] [SPARK-2024] renaming pickled -> pickledRDD
d998ad6 [Kan Zhang] [SPARK-2024] refectoring to get method params below 10
57a7a5e [Kan Zhang] [SPARK-2024] correcting typo
75ca5bd [Kan Zhang] [SPARK-2024] Better type checking for batch serialized RDD
0bdec55 [Kan Zhang] [SPARK-2024] Refactoring newly added tests
9f39ff4 [Kan Zhang] [SPARK-2024] Adding 2 saveAsHadoopDataset tests
0c134f3 [Kan Zhang] [SPARK-2024] Test refactoring and adding couple unbatched cases
7a176df [Kan Zhang] [SPARK-2024] Add saveAsSequenceFile to PySpark
This is a resubmission of #1452. It was reverted because it broke the build.
Currently (as of Spark 1.0.1), Spark sends RDD object (which contains closures) using Akka along with the task itself to the executors. This is inefficient because all tasks in the same stage use the same RDD object, but we have to send RDD object multiple times to the executors. This is especially bad when a closure references some variable that is very large. The current design led to users having to explicitly broadcast large variables.
The patch uses broadcast to send RDD objects and the closures to executors, and use Akka to only send a reference to the broadcast RDD/closure along with the partition specific information for the task. For those of you who know more about the internals, Spark already relies on broadcast to send the Hadoop JobConf every time it uses the Hadoop input, because the JobConf is large.
The user-facing impact of the change include:
1. Users won't need to decide what to broadcast anymore, unless they would want to use a large object multiple times in different operations
2. Task size will get smaller, resulting in faster scheduling and higher task dispatch throughput.
In addition, the change will simplify some internals of Spark, eliminating the need to maintain task caches and the complex logic to broadcast JobConf (which also led to a deadlock recently).
A simple way to test this:
```scala
val a = new Array[Byte](1000*1000); scala.util.Random.nextBytes(a);
sc.parallelize(1 to 1000, 1000).map { x => a; x }.groupBy { x => a; x }.count
```
Numbers on 3 r3.8xlarge instances on EC2
```
master branch: 5.648436068 s, 4.715361895 s, 5.360161877 s
with this change: 3.416348793 s, 1.477846558 s, 1.553432156 s
```
Author: Reynold Xin <rxin@apache.org>
Closes#1498 from rxin/broadcast-task and squashes the following commits:
f7364db [Reynold Xin] Code review feedback.
f8535dc [Reynold Xin] Fixed the style violation.
252238d [Reynold Xin] Serialize the final task closure as well as ShuffleDependency in taskBinary.
111007d [Reynold Xin] Fix broadcast tests.
797c247 [Reynold Xin] Properly send SparkListenerStageSubmitted and SparkListenerStageCompleted.
bab1d8b [Reynold Xin] Check for NotSerializableException in submitMissingTasks.
cf38450 [Reynold Xin] Use TorrentBroadcastFactory.
991c002 [Reynold Xin] Use HttpBroadcast.
de779f8 [Reynold Xin] Fix TaskContextSuite.
cc152fc [Reynold Xin] Don't cache the RDD broadcast variable.
d256b45 [Reynold Xin] Fixed unit test failures. One more to go.
cae0af3 [Reynold Xin] [SPARK-2521] Broadcast RDD object (instead of sending it along with every task).
Author: Koert Kuipers <koert@tresata.com>
Closes#735 from koertkuipers/feat-kryo-max-buffersize and squashes the following commits:
15f6d81 [Koert Kuipers] change default for spark.kryoserializer.buffer.max.mb to 64mb and add some documentation
1bcc22c [Koert Kuipers] Merge branch 'master' into feat-kryo-max-buffersize
0c9f8eb [Koert Kuipers] make default for kryo max buffer size 16MB
143ec4d [Koert Kuipers] test resizable buffer in kryo Output
0732445 [Koert Kuipers] support setting maxCapacity to something different than capacity in kryo Output
The current PR contains the following changes:
* Expose `DataType`s in the sql package (internal details are private to sql).
* Users can create Rows.
* Introduce `applySchema` to create a `SchemaRDD` by applying a `schema: StructType` to an `RDD[Row]`.
* Add a function `simpleString` to every `DataType`. Also, the schema represented by a `StructType` can be visualized by `printSchema`.
* `ScalaReflection.typeOfObject` provides a way to infer the Catalyst data type based on an object. Also, we can compose `typeOfObject` with some custom logics to form a new function to infer the data type (for different use cases).
* `JsonRDD` has been refactored to use changes introduced by this PR.
* Add a field `containsNull` to `ArrayType`. So, we can explicitly mark if an `ArrayType` can contain null values. The default value of `containsNull` is `false`.
New APIs are introduced in the sql package object and SQLContext. You can find the scaladoc at
[sql package object](http://yhuai.github.io/site/api/scala/index.html#org.apache.spark.sql.package) and [SQLContext](http://yhuai.github.io/site/api/scala/index.html#org.apache.spark.sql.SQLContext).
An example of using `applySchema` is shown below.
```scala
import org.apache.spark.sql._
val sqlContext = new org.apache.spark.sql.SQLContext(sc)
val schema =
StructType(
StructField("name", StringType, false) ::
StructField("age", IntegerType, true) :: Nil)
val people = sc.textFile("examples/src/main/resources/people.txt").map(_.split(",")).map(p => Row(p(0), p(1).trim.toInt))
val peopleSchemaRDD = sqlContext. applySchema(people, schema)
peopleSchemaRDD.printSchema
// root
// |-- name: string (nullable = false)
// |-- age: integer (nullable = true)
peopleSchemaRDD.registerAsTable("people")
sqlContext.sql("select name from people").collect.foreach(println)
```
I will add new contents to the SQL programming guide later.
JIRA: https://issues.apache.org/jira/browse/SPARK-2179
Author: Yin Huai <huai@cse.ohio-state.edu>
Closes#1346 from yhuai/dataTypeAndSchema and squashes the following commits:
1d45977 [Yin Huai] Clean up.
a6e08b4 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
c712fbf [Yin Huai] Converts types of values based on defined schema.
4ceeb66 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
e5f8df5 [Yin Huai] Scaladoc.
122d1e7 [Yin Huai] Address comments.
03bfd95 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
2476ed0 [Yin Huai] Minor updates.
ab71f21 [Yin Huai] Format.
fc2bed1 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
bd40a33 [Yin Huai] Address comments.
991f860 [Yin Huai] Move "asJavaDataType" and "asScalaDataType" to DataTypeConversions.scala.
1cb35fe [Yin Huai] Add "valueContainsNull" to MapType.
3edb3ae [Yin Huai] Python doc.
692c0b9 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
1d93395 [Yin Huai] Python APIs.
246da96 [Yin Huai] Add java data type APIs to javadoc index.
1db9531 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
d48fc7b [Yin Huai] Minor updates.
33c4fec [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
b9f3071 [Yin Huai] Java API for applySchema.
1c9f33c [Yin Huai] Java APIs for DataTypes and Row.
624765c [Yin Huai] Tests for applySchema.
aa92e84 [Yin Huai] Update data type tests.
8da1a17 [Yin Huai] Add Row.fromSeq.
9c99bc0 [Yin Huai] Several minor updates.
1d9c13a [Yin Huai] Update applySchema API.
85e9b51 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
e495e4e [Yin Huai] More comments.
42d47a3 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
c3f4a02 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
2e58dbd [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
b8b7db4 [Yin Huai] 1. Move sql package object and package-info to sql-core. 2. Minor updates on APIs. 3. Update scala doc.
68525a2 [Yin Huai] Update JSON unit test.
3209108 [Yin Huai] Add unit tests.
dcaf22f [Yin Huai] Add a field containsNull to ArrayType to indicate if an array can contain null values or not. If an ArrayType is constructed by "ArrayType(elementType)" (the existing constructor), the value of containsNull is false.
9168b83 [Yin Huai] Update comments.
fc649d7 [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
eca7d04 [Yin Huai] Add two apply methods which will be used to extract StructField(s) from a StructType.
949d6bb [Yin Huai] When creating a SchemaRDD for a JSON dataset, users can apply an existing schema.
7a6a7e5 [Yin Huai] Fix bug introduced by the change made on SQLContext.inferSchema.
43a45e1 [Yin Huai] Remove sql.util.package introduced in a previous commit.
0266761 [Yin Huai] Format
03eec4c [Yin Huai] Merge remote-tracking branch 'upstream/master' into dataTypeAndSchema
90460ac [Yin Huai] Infer the Catalyst data type from an object and cast a data value to the expected type.
3fa0df5 [Yin Huai] Provide easier ways to construct a StructType.
16be3e5 [Yin Huai] This commit contains three changes: * Expose `DataType`s in the sql package (internal details are private to sql). * Introduce `createSchemaRDD` to create a `SchemaRDD` from an `RDD` with a provided schema (represented by a `StructType`) and a provided function to construct `Row`, * Add a function `simpleString` to every `DataType`. Also, the schema represented by a `StructType` can be visualized by `printSchema`.
The main thing was that spark configs were not propagated to the driver, and so applications that do not specify `master` or `appName` automatically failed. This PR fixes that and a couple of miscellaneous things that are related.
One thing that may or may not be an issue is that the jars must be available on the driver node. In `standalone-cluster` mode, this effectively means these jars must be available on all the worker machines, since the driver is launched on one of them. The semantics here are not the same as `yarn-cluster` mode, where all the relevant jars are uploaded to a distributed cache automatically and shipped to the containers. This is probably not a concern, but still worth a mention.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1538 from andrewor14/standalone-cluster and squashes the following commits:
8c11a0d [Andrew Or] Clean up imports / comments (minor)
2678d13 [Andrew Or] Handle extraJavaOpts properly
7660547 [Andrew Or] Merge branch 'master' of github.com:apache/spark into standalone-cluster
6f64a9b [Andrew Or] Revert changes in YARN
2f2908b [Andrew Or] Fix tests
ed01491 [Andrew Or] Don't go overboard with escaping
8e105e1 [Andrew Or] Merge branch 'master' of github.com:apache/spark into standalone-cluster
b890949 [Andrew Or] Abstract usages of converting spark opts to java opts
79f63a3 [Andrew Or] Move sparkProps into javaOpts
78752f8 [Andrew Or] Fix tests
5a9c6c7 [Andrew Or] Fix line too long
c141a00 [Andrew Or] Don't display "unknown app" on driver log pages
d7e2728 [Andrew Or] Avoid deprecation warning in standalone Client
6ceb14f [Andrew Or] Allow relevant configs to propagate to standalone Driver
7f854bc [Andrew Or] Fix test
855256e [Andrew Or] Fix standalone-cluster mode
fd9da51 [Andrew Or] Formatting changes (minor)
As of Spark 1.0, RangePartitioner goes through data twice: once to compute the count and once to do sampling. As a result, to do sortByKey, Spark goes through data 3 times (once to count, once to sample, and once to sort).
`RangePartitioner` should go through data only once, collecting samples from input partitions as well as counting. If the data is balanced, this should give us a good sketch. If we see big partitions, we re-sample from them in order to collect enough items.
The downside is that we need to collect more from each partition in the first pass. An alternative solution is caching the intermediate result and decide whether to fetch the data after.
Author: Xiangrui Meng <meng@databricks.com>
Author: Reynold Xin <rxin@apache.org>
Closes#1562 from mengxr/range-partitioner and squashes the following commits:
6cc2551 [Xiangrui Meng] change foreach to for
eb39b08 [Xiangrui Meng] Merge branch 'master' into range-partitioner
eb95dd8 [Xiangrui Meng] separate sketching and determining bounds impl
c436d30 [Xiangrui Meng] fix binary metrics unit tests
db58a55 [Xiangrui Meng] add unit tests
a6e35d6 [Xiangrui Meng] minor update
60be09e [Xiangrui Meng] remove importance sampler
9ee9992 [Xiangrui Meng] update range partitioner to run only one job on roughly balanced data
cc12f47 [Xiangrui Meng] Merge remote-tracking branch 'apache/master' into range-part
06ac2ec [Xiangrui Meng] Merge remote-tracking branch 'apache/master' into range-part
17bcbf3 [Reynold Xin] Added seed.
badf20d [Reynold Xin] Renamed the method.
6940010 [Reynold Xin] Reservoir sampling implementation.
Author: Josh Rosen <joshrosen@apache.org>
Closes#1626 from JoshRosen/SPARK-2305 and squashes the following commits:
03fb283 [Josh Rosen] Update Py4J to version 0.8.2.1.
Implemented stratified sampling that guarantees exact sample size using ScaRSR with two passes over the RDD for sampling without replacement and three passes for sampling with replacement.
Author: Doris Xin <doris.s.xin@gmail.com>
Author: Xiangrui Meng <meng@databricks.com>
Closes#1025 from dorx/stratified and squashes the following commits:
245439e [Doris Xin] moved minSamplingRate to getUpperBound
eaf5771 [Doris Xin] bug fixes.
17a381b [Doris Xin] fixed a merge issue and a failed unit
ea7d27f [Doris Xin] merge master
b223529 [Xiangrui Meng] use approx bounds for poisson fix poisson mean for waitlisting add unit tests for Java
b3013a4 [Xiangrui Meng] move math3 back to test scope
eecee5f [Doris Xin] Merge branch 'master' into stratified
f4c21f3 [Doris Xin] Reviewer comments
a10e68d [Doris Xin] style fix
a2bf756 [Doris Xin] Merge branch 'master' into stratified
680b677 [Doris Xin] use mapPartitionWithIndex instead
9884a9f [Doris Xin] style fix
bbfb8c9 [Doris Xin] Merge branch 'master' into stratified
ee9d260 [Doris Xin] addressed reviewer comments
6b5b10b [Doris Xin] Merge branch 'master' into stratified
254e03c [Doris Xin] minor fixes and Java API.
4ad516b [Doris Xin] remove unused imports from PairRDDFunctions
bd9dc6e [Doris Xin] unit bug and style violation fixed
1fe1cff [Doris Xin] Changed fractionByKey to a map to enable arg check
944a10c [Doris Xin] [SPARK-2145] Add lower bound on sampling rate
0214a76 [Doris Xin] cleanUp
90d94c0 [Doris Xin] merge master
9e74ab5 [Doris Xin] Separated out most of the logic in sampleByKey
7327611 [Doris Xin] merge master
50581fc [Doris Xin] added a TODO for logging in python
46f6c8c [Doris Xin] fixed the NPE caused by closures being cleaned before being passed into the aggregate function
7e1a481 [Doris Xin] changed the permission on SamplingUtil
1d413ce [Doris Xin] fixed checkstyle issues
9ee94ee [Doris Xin] [SPARK-2082] stratified sampling in PairRDDFunctions that guarantees exact sample size
e3fd6a6 [Doris Xin] Merge branch 'master' into takeSample
7cab53a [Doris Xin] fixed import bug in rdd.py
ffea61a [Doris Xin] SPARK-1939: Refactor takeSample method in RDD
1441977 [Doris Xin] SPARK-1939 Refactor takeSample method in RDD to use ScaSRS
Datetime and time in Python will be converted into java.util.Calendar after serialization, it will be converted into java.sql.Timestamp during inferSchema().
In javaToPython(), Timestamp will be converted into Calendar, then be converted into datetime in Python after pickling.
Author: Davies Liu <davies.liu@gmail.com>
Closes#1601 from davies/date and squashes the following commits:
f0599b0 [Davies Liu] remove tests for sets and tuple in sql, fix list of list
c9d607a [Davies Liu] convert datetype for runtime
709d40d [Davies Liu] remove brackets
96db384 [Davies Liu] support datetime type for SchemaRDD
The pull request includes two changes:
1. Removes SortOrder introduced by SPARK-2125. The key ordering already includes the SortOrder information since an Ordering can be reverse. This is similar to Java's Comparator interface. Rarely does an API accept both a Comparator as well as a SortOrder.
2. Replaces the sortWith call in HashShuffleReader with an in-place quick sort.
Author: Reynold Xin <rxin@apache.org>
Closes#1631 from rxin/sortOrder and squashes the following commits:
c9d37e1 [Reynold Xin] [SPARK-2726] and [SPARK-2727] Remove SortOrder and do in-place sort.
Spark only transitively depends on the latter, based on the Hadoop version.
Author: Aaron Davidson <aaron@databricks.com>
Closes#1621 from aarondav/lang3 and squashes the following commits:
93c93bf [Aaron Davidson] Use commons-lang3 in SignalLogger rather than commons-lang
JIRA issue: [SPARK-2410](https://issues.apache.org/jira/browse/SPARK-2410)
Another try for #1399 & #1600. Those two PR breaks Jenkins builds because we made a separate profile `hive-thriftserver` in sub-project `assembly`, but the `hive-thriftserver` module is defined outside the `hive-thriftserver` profile. Thus every time a pull request that doesn't touch SQL code will also execute test suites defined in `hive-thriftserver`, but tests fail because related .class files are not included in the assembly jar.
In the most recent commit, module `hive-thriftserver` is moved into its own profile to fix this problem. All previous commits are squashed for clarity.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes#1620 from liancheng/jdbc-with-maven-fix and squashes the following commits:
629988e [Cheng Lian] Moved hive-thriftserver module definition into its own profile
ec3c7a7 [Cheng Lian] Cherry picked the Hive Thrift server
**Problem.** When caching, we currently unroll the entire RDD partition before making sure we have enough free memory. This is a common cause for OOMs especially when (1) the BlockManager has little free space left in memory, and (2) the partition is large.
**Solution.** We maintain a global memory pool of `M` bytes shared across all threads, similar to the way we currently manage memory for shuffle aggregation. Then, while we unroll each partition, periodically check if there is enough space to continue. If not, drop enough RDD blocks to ensure we have at least `M` bytes to work with, then try again. If we still don't have enough space to unroll the partition, give up and drop the block to disk directly if applicable.
**New configurations.**
- `spark.storage.bufferFraction` - the value of `M` as a fraction of the storage memory. (default: 0.2)
- `spark.storage.safetyFraction` - a margin of safety in case size estimation is slightly off. This is the equivalent of the existing `spark.shuffle.safetyFraction`. (default 0.9)
For more detail, see the [design document](https://issues.apache.org/jira/secure/attachment/12651793/spark-1777-design-doc.pdf). Tests pending for performance and memory usage patterns.
Author: Andrew Or <andrewor14@gmail.com>
Closes#1165 from andrewor14/them-rdd-memories and squashes the following commits:
e77f451 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
c7c8832 [Andrew Or] Simplify logic + update a few comments
269d07b [Andrew Or] Very minor changes to tests
6645a8a [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
b7e165c [Andrew Or] Add new tests for unrolling blocks
f12916d [Andrew Or] Slightly clean up tests
71672a7 [Andrew Or] Update unrollSafely tests
369ad07 [Andrew Or] Correct ensureFreeSpace and requestMemory behavior
f4d035c [Andrew Or] Allow one thread to unroll multiple blocks
a66fbd2 [Andrew Or] Rename a few things + update comments
68730b3 [Andrew Or] Fix weird scalatest behavior
e40c60d [Andrew Or] Fix MIMA excludes
ff77aa1 [Andrew Or] Fix tests
1a43c06 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
b9a6eee [Andrew Or] Simplify locking behavior on unrollMemoryMap
ed6cda4 [Andrew Or] Formatting fix (super minor)
f9ff82e [Andrew Or] putValues -> putIterator + putArray
beb368f [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
8448c9b [Andrew Or] Fix tests
a49ba4d [Andrew Or] Do not expose unroll memory check period
69bc0a5 [Andrew Or] Always synchronize on putLock before unrollMemoryMap
3f5a083 [Andrew Or] Simplify signature of ensureFreeSpace
dce55c8 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
8288228 [Andrew Or] Synchronize put and unroll properly
4f18a3d [Andrew Or] bufferFraction -> unrollFraction
28edfa3 [Andrew Or] Update a few comments / log messages
728323b [Andrew Or] Do not synchronize every 1000 elements
5ab2329 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
129c441 [Andrew Or] Fix bug: Use toArray rather than array
9a65245 [Andrew Or] Update a few comments + minor control flow changes
57f8d85 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
abeae4f [Andrew Or] Add comment clarifying the MEMORY_AND_DISK case
3dd96aa [Andrew Or] AppendOnlyBuffer -> Vector (+ a few small changes)
f920531 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
0871835 [Andrew Or] Add an effective storage level interface to BlockManager
64e7d4c [Andrew Or] Add/modify a few comments (minor)
8af2f35 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
4f4834e [Andrew Or] Use original storage level for blocks dropped to disk
ecc8c2d [Andrew Or] Fix binary incompatibility
24185ea [Andrew Or] Avoid dropping a block back to disk if reading from disk
2b7ee66 [Andrew Or] Fix bug in SizeTracking*
9b9a273 [Andrew Or] Fix tests
20eb3e5 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
649bdb3 [Andrew Or] Document spark.storage.bufferFraction
a10b0e7 [Andrew Or] Add initial memory request threshold + rename a few things
e9c3cb0 [Andrew Or] cacheMemoryMap -> unrollMemoryMap
198e374 [Andrew Or] Unfold -> unroll
0d50155 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
d9d02a8 [Andrew Or] Remove unused param in unfoldSafely
ec728d8 [Andrew Or] Add tests for safe unfolding of blocks
22b2209 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
078eb83 [Andrew Or] Add check for hasNext in PrimitiveVector.iterator
0871535 [Andrew Or] Fix tests in BlockManagerSuite
d68f31e [Andrew Or] Safely unfold blocks for all memory puts
5961f50 [Andrew Or] Fix tests
195abd7 [Andrew Or] Refactor: move unfold logic to MemoryStore
1e82d00 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
3ce413e [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
d5dd3b4 [Andrew Or] Free buffer memory in finally
ea02eec [Andrew Or] Fix tests
b8e1d9c [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
a8704c1 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
e1b8b25 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
87aa75c [Andrew Or] Fix mima excludes again (typo)
11eb921 [Andrew Or] Clarify comment (minor)
50cae44 [Andrew Or] Remove now duplicate mima exclude
7de5ef9 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
df47265 [Andrew Or] Fix binary incompatibility
6d05a81 [Andrew Or] Merge branch 'master' of github.com:apache/spark into them-rdd-memories
f94f5af [Andrew Or] Update a few comments (minor)
776aec9 [Andrew Or] Prevent OOM if a single RDD partition is too large
bbd3eea [Andrew Or] Fix CacheManagerSuite to use Array
97ea499 [Andrew Or] Change BlockManager interface to use Arrays
c12f093 [Andrew Or] Add SizeTrackingAppendOnlyBuffer and tests
(This is a replacement of #1399, trying to fix potential `HiveThriftServer2` port collision between parallel builds. Please refer to [these comments](https://github.com/apache/spark/pull/1399#issuecomment-50212572) for details.)
JIRA issue: [SPARK-2410](https://issues.apache.org/jira/browse/SPARK-2410)
Merging the Hive Thrift/JDBC server from [branch-1.0-jdbc](https://github.com/apache/spark/tree/branch-1.0-jdbc).
Thanks chenghao-intel for his initial contribution of the Spark SQL CLI.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes#1600 from liancheng/jdbc and squashes the following commits:
ac4618b [Cheng Lian] Uses random port for HiveThriftServer2 to avoid collision with parallel builds
090beea [Cheng Lian] Revert changes related to SPARK-2678, decided to move them to another PR
21c6cf4 [Cheng Lian] Updated Spark SQL programming guide docs
fe0af31 [Cheng Lian] Reordered spark-submit options in spark-shell[.cmd]
199e3fb [Cheng Lian] Disabled MIMA for hive-thriftserver
1083e9d [Cheng Lian] Fixed failed test suites
7db82a1 [Cheng Lian] Fixed spark-submit application options handling logic
9cc0f06 [Cheng Lian] Starts beeline with spark-submit
cfcf461 [Cheng Lian] Updated documents and build scripts for the newly added hive-thriftserver profile
061880f [Cheng Lian] Addressed all comments by @pwendell
7755062 [Cheng Lian] Adapts test suites to spark-submit settings
40bafef [Cheng Lian] Fixed more license header issues
e214aab [Cheng Lian] Added missing license headers
b8905ba [Cheng Lian] Fixed minor issues in spark-sql and start-thriftserver.sh
f975d22 [Cheng Lian] Updated docs for Hive compatibility and Shark migration guide draft
3ad4e75 [Cheng Lian] Starts spark-sql shell with spark-submit
a5310d1 [Cheng Lian] Make HiveThriftServer2 play well with spark-submit
61f39f4 [Cheng Lian] Starts Hive Thrift server via spark-submit
2c4c539 [Cheng Lian] Cherry picked the Hive Thrift server
This will decrease object allocation from the "update" closure used in map.changeValue.
Author: Matei Zaharia <matei@databricks.com>
Closes#1607 from mateiz/spark-2684 and squashes the following commits:
b7d89e6 [Matei Zaharia] Add insertAll for Iterables too, and fix some code style
561fc97 [Matei Zaharia] Update ExternalAppendOnlyMap to take an iterator as input
Author: Matei Zaharia <matei@databricks.com>
Closes#1593 from mateiz/spark-2680 and squashes the following commits:
3c949c4 [Matei Zaharia] Lower spark.shuffle.memoryFraction to 0.2 by default
Similar to SPARK-1034, the problem was that Py4J didn’t cope well with the fake ClassTags used in the Java API. It doesn’t look like there’s any reason why PythonRDD needs to take a ClassTag, since it just ignores the type of the previous RDD, so I removed the type parameter and we no longer pass ClassTags from Python.
Author: Josh Rosen <joshrosen@apache.org>
Closes#1605 from JoshRosen/spark-2601 and squashes the following commits:
b68e118 [Josh Rosen] Fix Py4J error when transforming pickleFiles [SPARK-2601]
handleMessageExecutor, handleReadWriteExecutor, and handleConnectExecutor are not marked as daemon and not named. I think there exists some condition in which Spark programs won't terminate because of this.
Stack dump attached in https://issues.apache.org/jira/browse/SPARK-2704
Author: Reynold Xin <rxin@apache.org>
Closes#1604 from rxin/daemon and squashes the following commits:
98d6a6c [Reynold Xin] [SPARK-2704] Name threads in ConnectionManager and mark them as daemon.
Added emptyRDD method to Java API with tests.
Author: bpaulin <bob@bobpaulin.com>
Closes#1597 from bobpaulin/SPARK-2279 and squashes the following commits:
5ad57c2 [bpaulin] [SPARK-2279] Added emptyRDD method to Java API
The current default value of spark.serializer.objectStreamReset is 10,000.
When trying to re-partition (e.g., to 64 partitions) a large file (e.g., 500MB), containing 1MB records, the serializer will cache 10000 x 1MB x 64 ~= 640 GB which will cause out of memory errors.
This patch sets the default value to a more reasonable default value (100).
Author: Hossein <hossein@databricks.com>
Closes#1595 from falaki/objectStreamReset and squashes the following commits:
650a935 [Hossein] Updated documentation
1aa0df8 [Hossein] Reduce default value of spark.serializer.objectStreamReset
Author: Josh Rosen <joshrosen@apache.org>
Closes#1596 from JoshRosen/spark-1458 and squashes the following commits:
fdbb0bf [Josh Rosen] Add SparkContext.version to Python & Java [SPARK-1458]
This is part of the scheduler cleanup/refactoring effort to make the scheduler code easier to maintain.
@kayousterhout @markhamstra please take a look ...
Author: Reynold Xin <rxin@apache.org>
Closes#1561 from rxin/dagSchedulerHashMaps and squashes the following commits:
1c44e15 [Reynold Xin] Clear pending tasks in submitMissingTasks.
620a0d1 [Reynold Xin] Use filterKeys.
5b54404 [Reynold Xin] Code review feedback.
c1e9a1c [Reynold Xin] Removed some HashMaps from DAGScheduler by storing information in Stage.
This reverts commit 06dc0d2c6b.
#1399 is making Jenkins fail. We should investigate and put this back after its passing tests.
Author: Michael Armbrust <michael@databricks.com>
Closes#1594 from marmbrus/revertJDBC and squashes the following commits:
59748da [Michael Armbrust] Revert "[SPARK-2410][SQL] Merging Hive Thrift/JDBC server"
Due to problems with when we update runningStages (in DAGScheduler.scala)
and how we decide to send a SparkListenerStageCompleted message to
SparkListeners, sometimes stages can be shown as "running" in the UI forever
(even after they have failed). This issue can manifest when stages are
resubmitted with 0 tasks, or when the DAGScheduler catches non-serializable
tasks. The problem also resulted in a (small) memory leak in the DAGScheduler,
where stages can stay in runningStages forever. This commit fixes
that problem and adds a unit test.
Thanks tsudukim for helping to look into this issue!
cc markhamstra rxin
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#1566 from kayousterhout/dag_fix and squashes the following commits:
217d74b [Kay Ousterhout] [SPARK-1726] [SPARK-2567] Eliminate zombie stages in UI.
This patch adds a sort flag into ShuffleDependecy and moves sort into hash shuffle implementation.
Moving sort into shuffle implementation can give space for other shuffle implementations (like sort-based shuffle) to better optimize sort through shuffle.
Author: jerryshao <saisai.shao@intel.com>
Closes#1210 from jerryshao/SPARK-2125 and squashes the following commits:
2feaf7b [jerryshao] revert MimaExcludes
ceddf75 [jerryshao] add MimaExeclude
f674ff4 [jerryshao] Add missing Scope restriction
b9fe0dd [jerryshao] Fix some style issues according to comments
ef6b729 [jerryshao] Change sort flag into Option
3f6eeed [jerryshao] Fix issues related to unit test
2f552a5 [jerryshao] Minor changes about naming and order
c92a281 [jerryshao] Move sort into shuffle implementations
JIRA issue:
- Main: [SPARK-2410](https://issues.apache.org/jira/browse/SPARK-2410)
- Related: [SPARK-2678](https://issues.apache.org/jira/browse/SPARK-2678)
Cherry picked the Hive Thrift/JDBC server from [branch-1.0-jdbc](https://github.com/apache/spark/tree/branch-1.0-jdbc).
(Thanks chenghao-intel for his initial contribution of the Spark SQL CLI.)
TODO
- [x] Use `spark-submit` to launch the server, the CLI and beeline
- [x] Migration guideline draft for Shark users
----
Hit by a bug in `SparkSubmitArguments` while working on this PR: all application options that are recognized by `SparkSubmitArguments` are stolen as `SparkSubmit` options. For example:
```bash
$ spark-submit --class org.apache.hive.beeline.BeeLine spark-internal --help
```
This actually shows usage information of `SparkSubmit` rather than `BeeLine`.
~~Fixed this bug here since the `spark-internal` related stuff also touches `SparkSubmitArguments` and I'd like to avoid conflict.~~
**UPDATE** The bug mentioned above is now tracked by [SPARK-2678](https://issues.apache.org/jira/browse/SPARK-2678). Decided to revert changes to this bug since it involves more subtle considerations and worth a separate PR.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes#1399 from liancheng/thriftserver and squashes the following commits:
090beea [Cheng Lian] Revert changes related to SPARK-2678, decided to move them to another PR
21c6cf4 [Cheng Lian] Updated Spark SQL programming guide docs
fe0af31 [Cheng Lian] Reordered spark-submit options in spark-shell[.cmd]
199e3fb [Cheng Lian] Disabled MIMA for hive-thriftserver
1083e9d [Cheng Lian] Fixed failed test suites
7db82a1 [Cheng Lian] Fixed spark-submit application options handling logic
9cc0f06 [Cheng Lian] Starts beeline with spark-submit
cfcf461 [Cheng Lian] Updated documents and build scripts for the newly added hive-thriftserver profile
061880f [Cheng Lian] Addressed all comments by @pwendell
7755062 [Cheng Lian] Adapts test suites to spark-submit settings
40bafef [Cheng Lian] Fixed more license header issues
e214aab [Cheng Lian] Added missing license headers
b8905ba [Cheng Lian] Fixed minor issues in spark-sql and start-thriftserver.sh
f975d22 [Cheng Lian] Updated docs for Hive compatibility and Shark migration guide draft
3ad4e75 [Cheng Lian] Starts spark-sql shell with spark-submit
a5310d1 [Cheng Lian] Make HiveThriftServer2 play well with spark-submit
61f39f4 [Cheng Lian] Starts Hive Thrift server via spark-submit
2c4c539 [Cheng Lian] Cherry picked the Hive Thrift server
Renaming `short` to `shortForm` and `long` to `longForm`.
JIRA: https://issues.apache.org/jira/browse/SPARK-2683
Author: Yin Huai <huai@cse.ohio-state.edu>
Closes#1585 from yhuai/SPARK-2683 and squashes the following commits:
5ddb843 [Yin Huai] "short" and "long" are Java keyworks. In order to generate javadoc, renaming "short" to "shortForm" and "long" to "longForm".
Author: Reynold Xin <rxin@apache.org>
Closes#1583 from rxin/closureClean and squashes the following commits:
8982fe6 [Reynold Xin] [SPARK-2529] Clean closures in foreach and foreachPartition.
JIRA: https://issues.apache.org/jira/browse/SPARK-2657
Our current code uses ArrayBuffers for each group of values in groupBy, as well as for the key's elements in CoGroupedRDD. ArrayBuffers have a lot of overhead if there are few values in them, which is likely to happen in cases such as join. In particular, they have a pointer to an Object[] of size 16 by default, which is 24 bytes for the array header + 128 for the pointers in there, plus at least 32 for the ArrayBuffer data structure. This patch replaces the per-group buffers with a CompactBuffer class that can store up to 2 elements more efficiently (in fields of itself) and acts like an ArrayBuffer beyond that. For a key's elements in CoGroupedRDD, we use an Array of CompactBuffers instead of an ArrayBuffer of ArrayBuffers.
There are some changes throughout the code to deal with CoGroupedRDD returning Array instead. We can also decide not to do that but CoGroupedRDD is a `DeveloperAPI` so I think it's okay to change it here.
Author: Matei Zaharia <matei@databricks.com>
Closes#1555 from mateiz/compact-groupby and squashes the following commits:
845a356 [Matei Zaharia] Lower initial size of CompactBuffer's vector to 8
07621a7 [Matei Zaharia] Review comments
0c1cd12 [Matei Zaharia] Don't use varargs in CompactBuffer.apply
bdc8a39 [Matei Zaharia] Small tweak to +=, and typos
f61f040 [Matei Zaharia] Fix line lengths
59da88b0 [Matei Zaharia] Fix line lengths
197cde8 [Matei Zaharia] Make CompactBuffer extend Seq to make its toSeq more efficient
775110f [Matei Zaharia] Change CoGroupedRDD to give (K, Array[Iterable[_]]) to avoid wrappers
9b4c6e8 [Matei Zaharia] Use CompactBuffer in CoGroupedRDD
ed577ab [Matei Zaharia] Use CompactBuffer in groupByKey
10f0de1 [Matei Zaharia] A CompactBuffer that's more memory-efficient than ArrayBuffer for small buffers
exact sample size not supported for now.
Author: Doris Xin <doris.s.xin@gmail.com>
Closes#1554 from dorx/pystratified and squashes the following commits:
4ba927a [Doris Xin] use rel diff (+- 50%) instead of abs diff (+- 50)
bdc3f8b [Doris Xin] updated unit to check sample holistically
7713c7b [Doris Xin] Python version of stratified sampling
During aggregation in Python worker, if the memory usage is above spark.executor.memory, it will do disk spilling aggregation.
It will split the aggregation into multiple stage, in each stage, it will partition the aggregated data by hash and dump them into disks. After all the data are aggregated, it will merge all the stages together (partition by partition).
Author: Davies Liu <davies.liu@gmail.com>
Closes#1460 from davies/spill and squashes the following commits:
cad91bf [Davies Liu] call gc.collect() after data.clear() to release memory as much as possible.
37d71f7 [Davies Liu] balance the partitions
902f036 [Davies Liu] add shuffle.py into run-tests
dcf03a9 [Davies Liu] fix memory_info() of psutil
67e6eba [Davies Liu] comment for MAX_TOTAL_PARTITIONS
f6bd5d6 [Davies Liu] rollback next_limit() again, the performance difference is huge:
e74b785 [Davies Liu] fix code style and change next_limit to memory_limit
400be01 [Davies Liu] address all the comments
6178844 [Davies Liu] refactor and improve docs
fdd0a49 [Davies Liu] add long doc string for ExternalMerger
1a97ce4 [Davies Liu] limit used memory and size of objects in partitionBy()
e6cc7f9 [Davies Liu] Merge branch 'master' into spill
3652583 [Davies Liu] address comments
e78a0a0 [Davies Liu] fix style
24cec6a [Davies Liu] get local directory by SPARK_LOCAL_DIR
57ee7ef [Davies Liu] update docs
286aaff [Davies Liu] let spilled aggregation in Python configurable
e9a40f6 [Davies Liu] recursive merger
6edbd1f [Davies Liu] Hash based disk spilling aggregation
Author: Neville Li <neville@spotify.com>
Closes#1188 from nevillelyh/neville/ui and squashes the following commits:
d3ac425 [Neville Li] SPARK-2250: show persisted RDD in stage UI
f075db9 [Neville Li] SPARK-2035: show call stack even when description is available
...rce manager UI
Use the event logger directory to provide a direct link to finished
application UI in yarn resourcemanager UI.
Author: Rahul Singhal <rahul.singhal@guavus.com>
Closes#1094 from rahulsinghaliitd/SPARK-2150 and squashes the following commits:
95f230c [Rahul Singhal] SPARK-2150: Provide direct link to finished application UI in yarn resource manager UI
...spark-submit
The PR allows invocations like
spark-submit --class org.MyClass --spark.shuffle.spill false myjar.jar
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1253 from sryza/sandy-spark-2310 and squashes the following commits:
1dc9855 [Sandy Ryza] More doc and cleanup
00edfb9 [Sandy Ryza] Review comments
91b244a [Sandy Ryza] Change format to --conf PROP=VALUE
8fabe77 [Sandy Ryza] SPARK-2310. Support arbitrary Spark properties on the command line with spark-submit
Author: GuoQiang Li <witgo@qq.com>
Closes#1511 from witgo/JsonProtocol and squashes the following commits:
2b6227f [GuoQiang Li] Fix NPE for JsonProtocol
Author: Ian O Connell <ioconnell@twitter.com>
Closes#1377 from ianoc/feature/SPARK-2102 and squashes the following commits:
5498566 [Ian O Connell] Docs update suggested by Patrick
20e8555 [Ian O Connell] Slight style change
f92c294 [Ian O Connell] Add docs for new KryoSerializer option
f3735c8 [Ian O Connell] Add using a kryo resource pool for the SqlSerializer
4e5c342 [Ian O Connell] Register the SparkConf for kryo, it gets swept into serialization
665805a [Ian O Connell] Add a spark.kryo.registrationRequired option for configuring the Kryo Serializer
Make spark's "local[N]" better.
In our company, we use "local[N]" in production. It works exellentlly. It's our best choice.
Author: woshilaiceshide <woshilaiceshide@qq.com>
Closes#1544 from woshilaiceshide/localX and squashes the following commits:
6c85154 [woshilaiceshide] [CORE] SPARK-2640: In "local[N]", free cores of the only executor should be touched by "spark.task.cpus" for every finish/start-up of tasks.
It's useful to know whether one thread is constantly spilling or multiple threads are spilling relatively infrequently. Right now everything looks a little jumbled and we can't tell which lines belong to the same thread. For instance:
```
06:14:37 ExternalAppendOnlyMap: Spilling in-memory map of 4 MB to disk (194 times so far)
06:14:37 ExternalAppendOnlyMap: Spilling in-memory map of 4 MB to disk (198 times so far)
06:14:37 ExternalAppendOnlyMap: Spilling in-memory map of 4 MB to disk (198 times so far)
06:14:37 ExternalAppendOnlyMap: Spilling in-memory map of 10 MB to disk (197 times so far)
06:14:38 ExternalAppendOnlyMap: Spilling in-memory map of 9 MB to disk (45 times so far)
06:14:38 ExternalAppendOnlyMap: Spilling in-memory map of 23 MB to disk (198 times so far)
06:14:38 ExternalAppendOnlyMap: Spilling in-memory map of 38 MB to disk (25 times so far)
06:14:38 ExternalAppendOnlyMap: Spilling in-memory map of 161 MB to disk (25 times so far)
06:14:39 ExternalAppendOnlyMap: Spilling in-memory map of 0 MB to disk (199 times so far)
06:14:39 ExternalAppendOnlyMap: Spilling in-memory map of 4 MB to disk (166 times so far)
06:14:39 ExternalAppendOnlyMap: Spilling in-memory map of 4 MB to disk (199 times so far)
06:14:39 ExternalAppendOnlyMap: Spilling in-memory map of 4 MB to disk (200 times so far)
```
Author: Andrew Or <andrewor14@gmail.com>
Closes#1517 from andrewor14/external-log and squashes the following commits:
90e48bb [Andrew Or] Log thread ID when spilling
The name `preservesPartitioning` is ambiguous: 1) preserves the indices of partitions, 2) preserves the partitioner. The latter is correct and `preservesPartitioning` should really be called `preservesPartitioner` to avoid confusion. Unfortunately, this is already part of the API and we cannot change. We should be clear in the doc and fix wrong usages.
This PR
1. adds notes in `maPartitions*`,
2. makes `RDD.sample` preserve partitioner,
3. changes `preservesPartitioning` to false in `RDD.zip` because the keys of the first RDD are no longer the keys of the zipped RDD,
4. fixes some wrong usages in MLlib.
Author: Xiangrui Meng <meng@databricks.com>
Closes#1526 from mengxr/preserve-partitioner and squashes the following commits:
b361e65 [Xiangrui Meng] update doc based on pwendell's comments
3b1ba19 [Xiangrui Meng] update doc
357575c [Xiangrui Meng] fix unit test
20b4816 [Xiangrui Meng] Merge branch 'master' into preserve-partitioner
d1caa65 [Xiangrui Meng] add doc to explain preservesPartitioning fix wrong usage of preservesPartitioning make sample preserse partitioning
The issue is caused by #1112 .
Author: GuoQiang Li <witgo@qq.com>
Closes#1501 from witgo/webui_style and squashes the following commits:
4b34998 [GuoQiang Li] In some cases, pages display incorrect in WebUI
### Why and what?
Currently, the AppendOnlyMap performs an "in-place" sort by converting its array of [key, value, key, value] pairs into a an array of [(key, value), (key, value)] pairs. However, this causes us to allocate many Tuple2 objects, which come at a nontrivial overhead.
This patch adds a Sorter API, intended for in memory sorts, which simply ports the Android Timsort implementation (available under Apache v2) and abstracts the interface in a way which introduces no more than 1 virtual function invocation of overhead at each abstraction point.
Please compare our port of the Android Timsort sort with the original implementation: http://www.diffchecker.com/wiwrykcl
### Memory implications
An AppendOnlyMap contains N kv pairs, which results in roughly 2N elements within its underlying array. Each of these elements is 4 bytes wide in a [compressed OOPS](https://wikis.oracle.com/display/HotSpotInternals/CompressedOops) system, which is the default.
Today's approach immediately allocates N Tuple2 objects, which take up 24N bytes in total (exposed via YourKit), and undergoes a Java sort. The Java 6 version immediately copies the entire array (4N bytes here), while the Java 7 version has a worst-case allocation of half the array (2N bytes).
This results in a worst-case sorting overhead of 24N + 2N = 26N bytes (for Java 7).
The Sorter does not require allocating any tuples, but since it uses Timsort, it may copy up to half the entire array in the worst case.
This results in a worst-case sorting overhead of 4N bytes.
Thus, we have reduced the worst-case overhead of the sort by roughly 22 bytes times the number of elements.
### Performance implications
As the destructiveSortedIterator is used for spilling in an ExternalAppendOnlyMap, the purpose of this patch is to provide stability by reducing memory usage rather than improve performance. However, because it implements Timsort, it also brings a substantial performance boost over our prior implementation.
Here are the results of a microbenchmark that sorted 25 million, randomly distributed (Float, Int) pairs. The Java Arrays.sort() tests were run **only on the keys**, and thus moved less data. Our current implementation is called "Tuple-sort using Arrays.sort()" while the new implementation is "KV-array using Sorter".
<table>
<tr><th>Test</th><th>First run (JDK6)</th><th>Average of 10 (JDK6)</th><th>First run (JDK7)</th><th>Average of 10 (JDK7)</th></tr>
<tr><td>primitive Arrays.sort()</td><td>3216 ms</td><td>1190 ms</td><td>2724 ms</td><td>131 ms (!!)</td></tr>
<tr><td>Arrays.sort()</td><td>18564 ms</td><td>2006 ms</td><td>13201 ms</td><td>878 ms</td></tr>
<tr><td>Tuple-sort using Arrays.sort()</td><td>31813 ms</td><td>3550 ms</td><td>20990 ms</td><td>1919 ms</td></tr>
<tr><td><b>KV-array using Sorter</b></td><td></td><td></td><td><b>15020 ms</b></td><td><b>834 ms</b></td></tr>
</table>
The results show that this Sorter performs exactly as expected (after the first run) -- it is as fast as the Java 7 Arrays.sort() (which shares the same algorithm), but is significantly faster than the Tuple-sort on Java 6 or 7.
In short, this patch should significantly improve performance for users running either Java 6 or 7.
Author: Aaron Davidson <aaron@databricks.com>
Closes#1502 from aarondav/sort and squashes the following commits:
652d936 [Aaron Davidson] Update license, move Sorter to java src
a7b5b1c [Aaron Davidson] fix licenses
5c0efaf [Aaron Davidson] Update tmpLength
ec395c8 [Aaron Davidson] Ignore benchmark (again) and fix docs
034bf10 [Aaron Davidson] Change to Apache v2 Timsort
b97296c [Aaron Davidson] Don't try to run benchmark on Jenkins + private[spark]
6307338 [Aaron Davidson] SPARK-2047: Introduce an in-mem Sorter, and use it to reduce mem usage
Changes RDD.toDebugString() to show hierarchy and shuffle transformations more clearly
New output:
```
(3) FlatMappedValuesRDD[325] at apply at Transformer.scala:22
| MappedValuesRDD[324] at apply at Transformer.scala:22
| CoGroupedRDD[323] at apply at Transformer.scala:22
+-(5) MappedRDD[320] at apply at Transformer.scala:22
| | MappedRDD[319] at apply at Transformer.scala:22
| | MappedValuesRDD[318] at apply at Transformer.scala:22
| | MapPartitionsRDD[317] at apply at Transformer.scala:22
| | ShuffledRDD[316] at apply at Transformer.scala:22
| +-(10) MappedRDD[315] at apply at Transformer.scala:22
| | ParallelCollectionRDD[314] at apply at Transformer.scala:22
+-(100) MappedRDD[322] at apply at Transformer.scala:22
| ParallelCollectionRDD[321] at apply at Transformer.scala:22
```
Author: Gregory Owen <greowen@gmail.com>
Closes#1364 from GregOwen/to-debug-string and squashes the following commits:
08f5c78 [Gregory Owen] toDebugString: prettier debug printing to show shuffles and joins more clearly
1603f7b [Gregory Owen] toDebugString: prettier debug printing to show shuffles and joins more clearly
As a result of shivaram's experience debugging long scheduler delay, I think we should improve the tooltip to point people in the right direction if scheduler delay is large.
Author: Kay Ousterhout <kayousterhout@gmail.com>
Closes#1488 from kayousterhout/better_tooltips and squashes the following commits:
22176fd [Kay Ousterhout] Improve scheduler delay tooltip.
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1474 from sryza/sandy-spark-2564 and squashes the following commits:
35b8388 [Sandy Ryza] Fix compile error on upmerge
7b985fb [Sandy Ryza] Fix test compile error
43f79e6 [Sandy Ryza] SPARK-2564. ShuffleReadMetrics.totalBlocksRead is redundant
We should fix this in branch-1.0 as well.
Author: Reynold Xin <rxin@apache.org>
Closes#1500 from rxin/rangePartitioner and squashes the following commits:
c0a94f5 [Reynold Xin] [SPARK-2598] RangePartitioner's binary search does not use the given Ordering.
...s of CoGroupedRDD and PairRDDFunctions
This also removes an unnecessary tuple creation in cogroup.
Author: Sandy Ryza <sandy@cloudera.com>
Closes#1447 from sryza/sandy-spark-2519-2 and squashes the following commits:
b6d9699 [Sandy Ryza] Remove missed Tuple2 match in CoGroupedRDD
a109828 [Sandy Ryza] Remove another pattern matching in MappedValuesRDD and revert some changes in PairRDDFunctions
be10f8a [Sandy Ryza] SPARK-2519 part 2. Remove pattern matching on Tuple2 in critical sections of CoGroupedRDD and PairRDDFunctions
This is a minor change. We should first logDebug($curRequestSize) and then set it to 0.
Author: Lijie Xu <csxulijie@gmail.com>
Closes#1477 from JerryLead/patch-1 and squashes the following commits:
aed722d [Lijie Xu] put 'curRequestSize = 0' after 'logDebug' it
Currently (as of Spark 1.0.1), Spark sends RDD object (which contains closures) using Akka along with the task itself to the executors. This is inefficient because all tasks in the same stage use the same RDD object, but we have to send RDD object multiple times to the executors. This is especially bad when a closure references some variable that is very large. The current design led to users having to explicitly broadcast large variables.
The patch uses broadcast to send RDD objects and the closures to executors, and use Akka to only send a reference to the broadcast RDD/closure along with the partition specific information for the task. For those of you who know more about the internals, Spark already relies on broadcast to send the Hadoop JobConf every time it uses the Hadoop input, because the JobConf is large.
The user-facing impact of the change include:
1. Users won't need to decide what to broadcast anymore, unless they would want to use a large object multiple times in different operations
2. Task size will get smaller, resulting in faster scheduling and higher task dispatch throughput.
In addition, the change will simplify some internals of Spark, eliminating the need to maintain task caches and the complex logic to broadcast JobConf (which also led to a deadlock recently).
A simple way to test this:
```scala
val a = new Array[Byte](1000*1000); scala.util.Random.nextBytes(a);
sc.parallelize(1 to 1000, 1000).map { x => a; x }.groupBy { x => a; x }.count
```
Numbers on 3 r3.8xlarge instances on EC2
```
master branch: 5.648436068 s, 4.715361895 s, 5.360161877 s
with this change: 3.416348793 s, 1.477846558 s, 1.553432156 s
```
Author: Reynold Xin <rxin@apache.org>
Closes#1452 from rxin/broadcast-task and squashes the following commits:
762e0be [Reynold Xin] Warn large broadcasts.
ade6eac [Reynold Xin] Log broadcast size.
c3b6f11 [Reynold Xin] Added a unit test for clean up.
754085f [Reynold Xin] Explain why broadcasting serialized copy of the task.
04b17f0 [Reynold Xin] [SPARK-2521] Broadcast RDD object once per TaskSet (instead of sending it for every task).
