Right now the spark shuffle service has a cache for index files. It is based on a # of files cached (spark.shuffle.service.index.cache.entries). This can cause issues if people have a lot of reducers because the size of each entry can fluctuate based on the # of reducers.
We saw an issues with a job that had 170000 reducers and it caused NM with spark shuffle service to use 700-800MB or memory in NM by itself.
We should change this cache to be memory based and only allow a certain memory size used. When I say memory based I mean the cache should have a limit of say 100MB.
https://issues.apache.org/jira/browse/SPARK-21501
Manual Testing with 170000 reducers has been performed with cache loaded up to max 100MB default limit, with each shuffle index file of size 1.3MB. Eviction takes place as soon as the total cache size reaches the 100MB limit and the objects will be ready for garbage collection there by avoiding NM to crash. No notable difference in runtime has been observed.
Author: Sanket Chintapalli <schintap@yahoo-inc.com>
Closes#18940 from redsanket/SPARK-21501.
## What changes were proposed in this pull request?
When we use `bin/spark-sql` command configuring `--conf spark.hadoop.foo=bar`, the `SparkSQLCliDriver` initializes an instance of hiveconf, it does not add `foo->bar` to it.
this pr gets `spark.hadoop.*` properties from sysProps to this hiveconf
## How was this patch tested?
UT
Author: hzyaoqin <hzyaoqin@corp.netease.com>
Author: Kent Yao <yaooqinn@hotmail.com>
Closes#18668 from yaooqinn/SPARK-21451.
## What changes were proposed in this pull request?
Fix 2 rendering errors on configuration doc page, due to SPARK-21243 and SPARK-15355.
## How was this patch tested?
Manually built and viewed docs with jekyll
Author: Sean Owen <sowen@cloudera.com>
Closes#18793 from srowen/SPARK-21593.
## What changes were proposed in this pull request?
Update the description of `spark.shuffle.maxChunksBeingTransferred` to include that the new coming connections will be closed when the max is hit and client should have retry mechanism.
Author: jinxing <jinxing6042@126.com>
Closes#18735 from jinxing64/SPARK-21530.
## What changes were proposed in this pull request?
A shuffle service can serves blocks from multiple apps/tasks. Thus the shuffle service can suffers high memory usage when lots of shuffle-reads happen at the same time. In my cluster, OOM always happens on shuffle service. Analyzing heap dump, memory cost by Netty(ChannelOutboundBufferEntry) can be up to 2~3G. It might make sense to reject "open blocks" request when memory usage is high on shuffle service.
93dd0c518d and 85c6ce6193 tried to alleviate the memory pressure on shuffle service but cannot solve the root cause. This pr proposes to control currency of shuffle read.
## How was this patch tested?
Added unit test.
Author: jinxing <jinxing6042@126.com>
Closes#18388 from jinxing64/SPARK-21175.
## What changes were proposed in this pull request?
For configurations with external shuffle enabled, we have observed that if a very large no. of blocks are being fetched from a remote host, it puts the NM under extra pressure and can crash it. This change introduces a configuration `spark.reducer.maxBlocksInFlightPerAddress` , to limit the no. of map outputs being fetched from a given remote address. The changes applied here are applicable for both the scenarios - when external shuffle is enabled as well as disabled.
## How was this patch tested?
Ran the job with the default configuration which does not change the existing behavior and ran it with few configurations of lower values -10,20,50,100. The job ran fine and there is no change in the output. (I will update the metrics related to NM in some time.)
Author: Dhruve Ashar <dhruveashar@gmail.com>
Closes#18487 from dhruve/impr/SPARK-21243.
## What changes were proposed in this pull request?
Spark provides several ways to set configurations, either from configuration file, or from `spark-submit` command line options, or programmatically through `SparkConf` class. It may confuses beginners why some configurations set through `SparkConf` cannot take affect. So here add some docs to address this problems and let beginners know how to correctly set configurations.
## How was this patch tested?
N/A
Author: jerryshao <sshao@hortonworks.com>
Closes#18552 from jerryshao/improve-doc.
## What changes were proposed in this pull request?
In current code, reducer can break the old shuffle service when `spark.reducer.maxReqSizeShuffleToMem` is enabled. Let's refine document.
Author: jinxing <jinxing6042@126.com>
Closes#18566 from jinxing64/SPARK-21343.
Current "--jars (spark.jars)", "--files (spark.files)", "--py-files (spark.submit.pyFiles)" and "--archives (spark.yarn.dist.archives)" only support non-glob path. This is OK for most of the cases, but when user requires to add more jars, files into Spark, it is too verbose to list one by one. So here propose to add glob path support for resources.
Also improving the code of downloading resources.
## How was this patch tested?
UT added, also verified manually in local cluster.
Author: jerryshao <sshao@hortonworks.com>
Closes#18235 from jerryshao/SPARK-21012.
## What changes were proposed in this pull request?
This change adds a new configuration option `spark.scheduler.listenerbus.eventqueue.size` to the configuration docs to specify the capacity of the spark listener bus event queue. Default value is 10000.
This is doc PR for [SPARK-15703](https://issues.apache.org/jira/browse/SPARK-15703).
I added option to the `Scheduling` section, however it might be more related to `Spark UI` section.
## How was this patch tested?
Manually verified correct rendering of configuration option.
Author: sadikovi <ivan.sadikov@lincolnuni.ac.nz>
Author: Ivan Sadikov <ivan.sadikov@team.telstra.com>
Closes#18476 from sadikovi/SPARK-20858.
## What changes were proposed in this pull request?
Disable spark.reducer.maxReqSizeShuffleToMem because it breaks the old shuffle service.
Credits to wangyum
Closes#18466
## How was this patch tested?
Jenkins
Author: Shixiong Zhu <shixiong@databricks.com>
Author: Yuming Wang <wgyumg@gmail.com>
Closes#18467 from zsxwing/SPARK-21253.
## What changes were proposed in this pull request?
Currently we are running into an issue with Yarn work preserving enabled + external shuffle service.
In the work preserving enabled scenario, the failure of NM will not lead to the exit of executors, so executors can still accept and run the tasks. The problem here is when NM is failed, external shuffle service is actually inaccessible, so reduce tasks will always complain about the “Fetch failure”, and the failure of reduce stage will make the parent stage (map stage) rerun. The tricky thing here is Spark scheduler is not aware of the unavailability of external shuffle service, and will reschedule the map tasks on the executor where NM is failed, and again reduce stage will be failed with “Fetch failure”, and after 4 retries, the job is failed. This could also apply to other cluster manager with external shuffle service.
So here the main problem is that we should avoid assigning tasks to those bad executors (where shuffle service is unavailable). Current Spark's blacklist mechanism could blacklist executors/nodes by failure tasks, but it doesn't handle this specific fetch failure scenario. So here propose to improve the current application blacklist mechanism to handle fetch failure issue (especially with external shuffle service unavailable issue), to blacklist the executors/nodes where shuffle fetch is unavailable.
## How was this patch tested?
Unit test and small cluster verification.
Author: jerryshao <sshao@hortonworks.com>
Closes#17113 from jerryshao/SPARK-13669.
## What changes were proposed in this pull request?
Add lost `<tr>` tag for `configuration.md`.
## How was this patch tested?
N/A
Author: Yuming Wang <wgyumg@gmail.com>
Closes#18372 from wangyum/docs-missing-tr.
## What changes were proposed in this pull request?
Currently the shuffle service registration timeout and retry has been hardcoded. This works well for small workloads but under heavy workload when the shuffle service is busy transferring large amount of data we see significant delay in responding to the registration request, as a result we often see the executors fail to register with the shuffle service, eventually failing the job. We need to make these two parameters configurable.
## How was this patch tested?
* Updated `BlockManagerSuite` to test registration timeout and max attempts configuration actually works.
cc sitalkedia
Author: Li Yichao <lyc@zhihu.com>
Closes#18092 from liyichao/SPARK-20640.
## What changes were proposed in this pull request?
In our use case of launching Spark applications via REST APIs (Livy), there's no way for user to specify command line arguments, all Spark configurations are set through configurations map. For "--repositories" because there's no equivalent Spark configuration, so we cannot specify the custom repository through configuration.
So here propose to add "--repositories" equivalent configuration in Spark.
## How was this patch tested?
New UT added.
Author: jerryshao <sshao@hortonworks.com>
Closes#18201 from jerryshao/SPARK-20981.
## What changes were proposed in this pull request?
Currently the whole block is fetched into memory(off heap by default) when shuffle-read. A block is defined by (shuffleId, mapId, reduceId). Thus it can be large when skew situations. If OOM happens during shuffle read, job will be killed and users will be notified to "Consider boosting spark.yarn.executor.memoryOverhead". Adjusting parameter and allocating more memory can resolve the OOM. However the approach is not perfectly suitable for production environment, especially for data warehouse.
Using Spark SQL as data engine in warehouse, users hope to have a unified parameter(e.g. memory) but less resource wasted(resource is allocated but not used). The hope is strong especially when migrating data engine to Spark from another one(e.g. Hive). Tuning the parameter for thousands of SQLs one by one is very time consuming.
It's not always easy to predict skew situations, when happen, it make sense to fetch remote blocks to disk for shuffle-read, rather than kill the job because of OOM.
In this pr, I propose to fetch big blocks to disk(which is also mentioned in SPARK-3019):
1. Track average size and also the outliers(which are larger than 2*avgSize) in MapStatus;
2. Request memory from `MemoryManager` before fetch blocks and release the memory to `MemoryManager` when `ManagedBuffer` is released.
3. Fetch remote blocks to disk when failing acquiring memory from `MemoryManager`, otherwise fetch to memory.
This is an improvement for memory control when shuffle blocks and help to avoid OOM in scenarios like below:
1. Single huge block;
2. Sizes of many blocks are underestimated in `MapStatus` and the actual footprint of blocks is much larger than the estimated.
## How was this patch tested?
Added unit test in `MapStatusSuite` and `ShuffleBlockFetcherIteratorSuite`.
Author: jinxing <jinxing6042@126.com>
Closes#16989 from jinxing64/SPARK-19659.
## What changes were proposed in this pull request?
Currently, when number of reduces is above 2000, HighlyCompressedMapStatus is used to store size of blocks. in HighlyCompressedMapStatus, only average size is stored for non empty blocks. Which is not good for memory control when we shuffle blocks. It makes sense to store the accurate size of block when it's above threshold.
## How was this patch tested?
Added test in MapStatusSuite.
Author: jinxing <jinxing6042@126.com>
Closes#18031 from jinxing64/SPARK-20801.
This change does a more thorough redaction of sensitive information from logs and UI
Add unit tests that ensure that no regressions happen that leak sensitive information to the logs.
The motivation for this change was appearance of password like so in `SparkListenerEnvironmentUpdate` in event logs under some JVM configurations:
`"sun.java.command":"org.apache.spark.deploy.SparkSubmit ... --conf spark.executorEnv.HADOOP_CREDSTORE_PASSWORD=secret_password ..."
`
Previously redaction logic was only checking if the key matched the secret regex pattern, it'd redact it's value. That worked for most cases. However, in the above case, the key (sun.java.command) doesn't tell much, so the value needs to be searched. This PR expands the check to check for values as well.
## How was this patch tested?
New unit tests added that ensure that no sensitive information is present in the event logs or the yarn logs. Old unit test in UtilsSuite was modified because the test was asserting that a non-sensitive property's value won't be redacted. However, the non-sensitive value had the literal "secret" in it which was causing it to redact. Simply updating the non-sensitive property's value to another arbitrary value (that didn't have "secret" in it) fixed it.
Author: Mark Grover <mark@apache.org>
Closes#17725 from markgrover/spark-20435.
## What changes were proposed in this pull request?
Simple documentation change to remove explicit vendor references.
## How was this patch tested?
NA
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: anabranch <bill@databricks.com>
Closes#17695 from anabranch/remove-vendor.
## What changes were proposed in this pull request?
Pregel-based iterative algorithms with more than ~50 iterations begin to slow down and eventually fail with a StackOverflowError due to Spark's lack of support for long lineage chains.
This PR causes Pregel to checkpoint the graph periodically if the checkpoint directory is set.
This PR moves PeriodicGraphCheckpointer.scala from mllib to graphx, moves PeriodicRDDCheckpointer.scala, PeriodicCheckpointer.scala from mllib to core
## How was this patch tested?
unit tests, manual tests
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
(If this patch involves UI changes, please attach a screenshot; otherwise, remove this)
Author: ding <ding@localhost.localdomain>
Author: dding3 <ding.ding@intel.com>
Author: Michael Allman <michael@videoamp.com>
Closes#15125 from dding3/cp2_pregel.
## What changes were proposed in this pull request?
Use the REST interface submits the spark job.
e.g.
curl -X POST http://10.43.183.120:6066/v1/submissions/create --header "Content-Type:application/json;charset=UTF-8" --data'{
"action": "CreateSubmissionRequest",
"appArgs": [
"myAppArgument"
],
"appResource": "/home/mr/gxl/test.jar",
"clientSparkVersion": "2.2.0",
"environmentVariables": {
"SPARK_ENV_LOADED": "1"
},
"mainClass": "cn.zte.HdfsTest",
"sparkProperties": {
"spark.jars": "/home/mr/gxl/test.jar",
**"spark.driver.supervise": "true",**
"spark.app.name": "HdfsTest",
"spark.eventLog.enabled": "false",
"spark.submit.deployMode": "cluster",
"spark.master": "spark://10.43.183.120:6066"
}
}'
**I hope that make sure that the driver is automatically restarted if it fails with non-zero exit code.
But I can not find the 'spark.driver.supervise' configuration parameter specification and default values from the spark official document.**
## How was this patch tested?
manual tests
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: 郭小龙 10207633 <guo.xiaolong1@zte.com.cn>
Author: guoxiaolong <guo.xiaolong1@zte.com.cn>
Author: guoxiaolongzte <guo.xiaolong1@zte.com.cn>
Closes#17696 from guoxiaolongzte/SPARK-20401.
…anges.
## What changes were proposed in this pull request?
Document compression way little detail changes.
1.spark.eventLog.compress add 'Compression will use spark.io.compression.codec.'
2.spark.broadcast.compress add 'Compression will use spark.io.compression.codec.'
3,spark.rdd.compress add 'Compression will use spark.io.compression.codec.'
4.spark.io.compression.codec add 'event log describe'.
eg
Through the documents, I don't know what is compression mode about 'event log'.
## How was this patch tested?
manual tests
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: 郭小龙 10207633 <guo.xiaolong1@zte.com.cn>
Closes#17498 from guoxiaolongzte/SPARK-20177.
The previously hardcoded max 4 retries per stage is not suitable for all cluster configurations. Since spark retries a stage at the sign of the first fetch failure, you can easily end up with many stage retries to discover all the failures. In particular, two scenarios this value should change are (1) if there are more than 4 executors per node; in that case, it may take 4 retries to discover the problem with each executor on the node and (2) during cluster maintenance on large clusters, where multiple machines are serviced at once, but you also cannot afford total cluster downtime. By making this value configurable, cluster managers can tune this value to something more appropriate to their cluster configuration.
Unit tests
Author: Sital Kedia <skedia@fb.com>
Closes#17307 from sitalkedia/SPARK-13369.
## What changes were proposed in this pull request?
We are proposing addition of pro-active block replication in case of executor failures. BlockManagerMasterEndpoint does all the book-keeping to keep a track of all the executors and the blocks they hold. It also keeps a track of which executors are alive through heartbeats. When an executor is removed, all this book-keeping state is updated to reflect the lost executor. This step can be used to identify executors that are still in possession of a copy of the cached data and a message could be sent to them to use the existing "replicate" function to find and place new replicas on other suitable hosts. Blocks replicated this way will let the master know of their existence.
This can happen when an executor is lost, and would that way be pro-active as opposed be being done at query time.
## How was this patch tested?
This patch was tested with existing unit tests along with new unit tests added to test the functionality.
Author: Shubham Chopra <schopra31@bloomberg.net>
Closes#14412 from shubhamchopra/ProactiveBlockReplication.
## What changes were proposed in this pull request?
In SPARK-8425, we introduced a mechanism for blacklisting executors and nodes (hosts). After a certain number of failures, these resources would be "blacklisted" and no further work would be assigned to them for some period of time.
In some scenarios, it is better to fail fast, and to simply kill these unreliable resources. This changes proposes to do so by having the BlacklistTracker kill unreliable resources when they would otherwise be "blacklisted".
In order to be thread safe, this code depends on the CoarseGrainedSchedulerBackend sending a message to the driver backend in order to do the actual killing. This also helps to prevent a race which would permit work to begin on a resource (executor or node), between the time the resource is marked for killing and the time at which it is finally killed.
## How was this patch tested?
./dev/run-tests
Ran https://github.com/jsoltren/jose-utils/blob/master/blacklist/test-blacklist.sh, and checked logs to see executors and nodes being killed.
Testing can likely be improved here; suggestions welcome.
Author: José Hiram Soltren <jose@cloudera.com>
Closes#16650 from jsoltren/SPARK-16554-submit.
Make the SSL port configuration explicit, instead of deriving it
from the non-SSL port, but retain the existing functionality in
case anyone depends on it.
The change starts the HTTPS and HTTP connectors separately, so
that it's possible to use independent ports for each. For that to
work, the initialization of the server needs to be shuffled around
a bit. The change also makes it so the initialization of both
connectors is similar, and end up using the same Scheduler - previously
only the HTTP connector would use the correct one.
Also fixed some outdated documentation about a couple of services
that were removed long ago.
Tested with unit tests and by running spark-shell with SSL configs.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#16625 from vanzin/SPARK-17874.
This change introduces a new auth mechanism to the transport library,
to be used when users enable strong encryption. This auth mechanism
has better security than the currently used DIGEST-MD5.
The new protocol uses symmetric key encryption to mutually authenticate
the endpoints, and is very loosely based on ISO/IEC 9798.
The new protocol falls back to SASL when it thinks the remote end is old.
Because SASL does not support asking the server for multiple auth protocols,
which would mean we could re-use the existing SASL code by just adding a
new SASL provider, the protocol is implemented outside of the SASL API
to avoid the boilerplate of adding a new provider.
Details of the auth protocol are discussed in the included README.md
file.
This change partly undos the changes added in SPARK-13331; AES encryption
is now decoupled from SASL authentication. The encryption code itself,
though, has been re-used as part of this change.
## How was this patch tested?
- Unit tests
- Tested Spark 2.2 against Spark 1.6 shuffle service with SASL enabled
- Tested Spark 2.2 against Spark 2.2 shuffle service with SASL fallback disabled
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#16521 from vanzin/SPARK-19139.
## What changes were proposed in this pull request?
Fix typo in docs
## How was this patch tested?
Author: uncleGen <hustyugm@gmail.com>
Closes#16658 from uncleGen/typo-issue.
## What changes were proposed in this pull request?
Drop more elements when `stageData.taskData.size > retainedTasks` to reduce the number of times on call drop function.
## How was this patch tested?
Jenkins
Author: Yuming Wang <wgyumg@gmail.com>
Closes#16527 from wangyum/SPARK-19146.
## What changes were proposed in this pull request?
Adding option in spark-submit to allow overriding the default IvySettings used to resolve artifacts as part of the Spark Packages functionality. This will allow all artifact resolution to go through a central managed repository, such as Nexus or Artifactory, where site admins can better approve and control what is used with Spark apps.
This change restructures the creation of the IvySettings object in two distinct ways. First, if the `spark.ivy.settings` option is not defined then `buildIvySettings` will create a default settings instance, as before, with defined repositories (Maven Central) included. Second, if the option is defined, the ivy settings file will be loaded from the given path and only repositories defined within will be used for artifact resolution.
## How was this patch tested?
Existing tests for default behaviour, Manual tests that load a ivysettings.xml file with local and Nexus repositories defined. Added new test to load a simple Ivy settings file with a local filesystem resolver.
Author: Bryan Cutler <cutlerb@gmail.com>
Author: Ian Hummel <ian@themodernlife.net>
Closes#15119 from BryanCutler/spark-custom-IvySettings.
## What changes were proposed in this pull request?
configuration.html section headings were not specified correctly in markdown and weren't rendering, being recognized correctly. Removed extra p tags and pulled level 4 titles up to level 3, since level 3 had been skipped. This improves the TOC.
## How was this patch tested?
Doc build, manual check.
Author: Sean Owen <sowen@cloudera.com>
Closes#16490 from srowen/SPARK-19106.
## What changes were proposed in this pull request?
On configuration doc page:https://spark.apache.org/docs/latest/configuration.html
We mentioned spark.kryoserializer.buffer.max : Maximum allowable size of Kryo serialization buffer. This must be larger than any object you attempt to serialize. Increase this if you get a "buffer limit exceeded" exception inside Kryo.
from source code, it has hard coded upper limit :
```
val maxBufferSizeMb = conf.getSizeAsMb("spark.kryoserializer.buffer.max", "64m").toInt
if (maxBufferSizeMb >= ByteUnit.GiB.toMiB(2))
{ throw new IllegalArgumentException("spark.kryoserializer.buffer.max must be less than " + s"2048 mb, got: + $maxBufferSizeMb mb.") }
```
We should mention "this value must be less than 2048 mb" on the configuration doc page as well.
## How was this patch tested?
None. Since it's minor doc change.
Author: Yuexin Zhang <yxzhang@cloudera.com>
Closes#16412 from cnZach/SPARK-19006.
## What changes were proposed in this pull request?
Spark's current task cancellation / task killing mechanism is "best effort" because some tasks may not be interruptible or may not respond to their "killed" flags being set. If a significant fraction of a cluster's task slots are occupied by tasks that have been marked as killed but remain running then this can lead to a situation where new jobs and tasks are starved of resources that are being used by these zombie tasks.
This patch aims to address this problem by adding a "task reaper" mechanism to executors. At a high-level, task killing now launches a new thread which attempts to kill the task and then watches the task and periodically checks whether it has been killed. The TaskReaper will periodically re-attempt to call `TaskRunner.kill()` and will log warnings if the task keeps running. I modified TaskRunner to rename its thread at the start of the task, allowing TaskReaper to take a thread dump and filter it in order to log stacktraces from the exact task thread that we are waiting to finish. If the task has not stopped after a configurable timeout then the TaskReaper will throw an exception to trigger executor JVM death, thereby forcibly freeing any resources consumed by the zombie tasks.
This feature is flagged off by default and is controlled by four new configurations under the `spark.task.reaper.*` namespace. See the updated `configuration.md` doc for details.
## How was this patch tested?
Tested via a new test case in `JobCancellationSuite`, plus manual testing.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#16189 from JoshRosen/cancellation.
## What changes were proposed in this pull request?
This builds upon the blacklisting introduced in SPARK-17675 to add blacklisting of executors and nodes for an entire Spark application. Resources are blacklisted based on tasks that fail, in tasksets that eventually complete successfully; they are automatically returned to the pool of active resources based on a timeout. Full details are available in a design doc attached to the jira.
## How was this patch tested?
Added unit tests, ran them via Jenkins, also ran a handful of them in a loop to check for flakiness.
The added tests include:
- verifying BlacklistTracker works correctly
- verifying TaskSchedulerImpl interacts with BlacklistTracker correctly (via a mock BlacklistTracker)
- an integration test for the entire scheduler with blacklisting in a few different scenarios
Author: Imran Rashid <irashid@cloudera.com>
Author: mwws <wei.mao@intel.com>
Closes#14079 from squito/blacklist-SPARK-8425.
This change moves the logic that translates Spark configuration to
commons-crypto configuration to the network-common module. It also
extends TransportConf and ConfigProvider to provide the necessary
interfaces for the translation to work.
As part of the change, I removed SystemPropertyConfigProvider, which
was mostly used as an "empty config" in unit tests, and adjusted the
very few tests that required a specific config.
I also changed the config keys for AES encryption to live under the
"spark.network." namespace, which is more correct than their previous
names under "spark.authenticate.".
Tested via existing unit test.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#16200 from vanzin/SPARK-18773.
This PR adds `spark.ui.showConsoleProgress` to the configuration docs.
I tested this PR by building the docs locally and confirming that this change shows up as expected.
Relates to #3029.
Author: Nicholas Chammas <nicholas.chammas@gmail.com>
Closes#16151 from nchammas/ui-progressbar-doc.
This change modifies the method used to propagate encryption keys used during
shuffle. Instead of relying on YARN's UserGroupInformation credential propagation,
this change explicitly distributes the key using the messages exchanged between
driver and executor during registration. When RPC encryption is enabled, this means
key propagation is also secure.
This allows shuffle encryption to work in non-YARN mode, which means that it's
easier to write unit tests for areas of the code that are affected by the feature.
The key is stored in the SecurityManager; because there are many instances of
that class used in the code, the key is only guaranteed to exist in the instance
managed by the SparkEnv. This path was chosen to avoid storing the key in the
SparkConf, which would risk having the key being written to disk as part of the
configuration (as, for example, is done when starting YARN applications).
Tested by new and existing unit tests (which were moved from the YARN module to
core), and by running apps with shuffle encryption enabled.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#15981 from vanzin/SPARK-18547.
## What changes were proposed in this pull request?
This patch adds a new property called `spark.secret.redactionPattern` that
allows users to specify a scala regex to decide which Spark configuration
properties and environment variables in driver and executor environments
contain sensitive information. When this regex matches the property or
environment variable name, its value is redacted from the environment UI and
various logs like YARN and event logs.
This change uses this property to redact information from event logs and YARN
logs. It also, updates the UI code to adhere to this property instead of
hardcoding the logic to decipher which properties are sensitive.
Here's an image of the UI post-redaction:
Here's the text in the YARN logs, post-redaction:
``HADOOP_CREDSTORE_PASSWORD -> *********(redacted)``
Here's the text in the event logs, post-redaction:
``...,"spark.executorEnv.HADOOP_CREDSTORE_PASSWORD":"*********(redacted)","spark.yarn.appMasterEnv.HADOOP_CREDSTORE_PASSWORD":"*********(redacted)",...``
## How was this patch tested?
1. Unit tests are added to ensure that redaction works.
2. A YARN job reading data off of S3 with confidential information
(hadoop credential provider password) being provided in the environment
variables of driver and executor. And, afterwards, logs were grepped to make
sure that no mention of secret password was present. It was also ensure that
the job was able to read the data off of S3 correctly, thereby ensuring that
the sensitive information was being trickled down to the right places to read
the data.
3. The event logs were checked to make sure no mention of secret password was
present.
4. UI environment tab was checked to make sure there was no secret information
being displayed.
Author: Mark Grover <mark@apache.org>
Closes#15971 from markgrover/master_redaction.
## What changes were proposed in this pull request?
Avoid hard-coding spark.rpc.askTimeout to non-default in Client; fix doc about spark.rpc.askTimeout default
## How was this patch tested?
Existing tests
Author: Sean Owen <sowen@cloudera.com>
Closes#15833 from srowen/SPARK-18353.
## What changes were proposed in this pull request?
Fix typos in the 'configuration', 'monitoring' and 'sql-programming-guide' documentation.
## How was this patch tested?
Manually.
Author: Weiqing Yang <yangweiqing001@gmail.com>
Closes#15886 from weiqingy/fixTypo.
## What changes were proposed in this pull request?
Many applications take Spark as a computing engine and run on it. This PR adds a configuration property `spark.log.callerContext` that can be used by Spark's upstream applications (e.g. Oozie) to set up their caller contexts into Spark. In the end, Spark will combine its own caller context with the caller contexts of its upstream applications, and write them into Yarn RM log and HDFS audit log.
The audit log has a config to truncate the caller contexts passed in (default 128). The caller contexts will be sent over rpc, so it should be concise. The call context written into HDFS log and Yarn log consists of two parts: the information `A` specified by Spark itself and the value `B` of `spark.log.callerContext` property. Currently `A` typically takes 64 to 74 characters, so `B` can have up to 50 characters (mentioned in the doc `running-on-yarn.md`)
## How was this patch tested?
Manual tests. I have run some Spark applications with `spark.log.callerContext` configuration in Yarn client/cluster mode, and verified that the caller contexts were written into Yarn RM log and HDFS audit log correctly.
The ways to configure `spark.log.callerContext` property:
- In spark-defaults.conf:
```
spark.log.callerContext infoSpecifiedByUpstreamApp
```
- In app's source code:
```
val spark = SparkSession
.builder
.appName("SparkKMeans")
.config("spark.log.callerContext", "infoSpecifiedByUpstreamApp")
.getOrCreate()
```
When running on Spark Yarn cluster mode, the driver is unable to pass 'spark.log.callerContext' to Yarn client and AM since Yarn client and AM have already started before the driver performs `.config("spark.log.callerContext", "infoSpecifiedByUpstreamApp")`.
The following example shows the command line used to submit a SparkKMeans application and the corresponding records in Yarn RM log and HDFS audit log.
Command:
```
./bin/spark-submit --verbose --executor-cores 3 --num-executors 1 --master yarn --deploy-mode client --class org.apache.spark.examples.SparkKMeans examples/target/original-spark-examples_2.11-2.1.0-SNAPSHOT.jar hdfs://localhost:9000/lr_big.txt 2 5
```
Yarn RM log:
<img width="1440" alt="screen shot 2016-10-19 at 9 12 03 pm" src="https://cloud.githubusercontent.com/assets/8546874/19547050/7d2f278c-9649-11e6-9df8-8d5ff12609f0.png">
HDFS audit log:
<img width="1400" alt="screen shot 2016-10-19 at 10 18 14 pm" src="https://cloud.githubusercontent.com/assets/8546874/19547102/096060ae-964a-11e6-981a-cb28efd5a058.png">
Author: Weiqing Yang <yangweiqing001@gmail.com>
Closes#15563 from weiqingy/SPARK-16759.
## What changes were proposed in this pull request?
DIGEST-MD5 mechanism is used for SASL authentication and secure communication. DIGEST-MD5 mechanism supports 3DES, DES, and RC4 ciphers. However, 3DES, DES and RC4 are slow relatively.
AES provide better performance and security by design and is a replacement for 3DES according to NIST. Apache Common Crypto is a cryptographic library optimized with AES-NI, this patch employ Apache Common Crypto as enc/dec backend for SASL authentication and secure channel to improve spark RPC.
## How was this patch tested?
Unit tests and Integration test.
Author: Junjie Chen <junjie.j.chen@intel.com>
Closes#15172 from cjjnjust/shuffle_rpc_encrypt.
## What changes were proposed in this pull request?
This Pull request comprises of the critical bug SPARK-16575 changes. This change rectifies the issue with BinaryFileRDD partition calculations as upon creating an RDD with sc.binaryFiles, the resulting RDD always just consisted of two partitions only.
## How was this patch tested?
The original issue ie. getNumPartitions on binary Files RDD (always having two partitions) was first replicated and then tested upon the changes. Also the unit tests have been checked and passed.
This contribution is my original work and I licence the work to the project under the project's open source license
srowen hvanhovell rxin vanzin skyluc kmader zsxwing datafarmer Please have a look .
Author: fidato <fidato.july13@gmail.com>
Closes#15327 from fidato13/SPARK-16575.
In SPARK-4761 / #3621 (December 2014) we enabled Kryo serialization by default in the Spark Thrift Server. However, I don't think that the original rationale for doing this still holds now that most Spark SQL serialization is now performed via encoders and our UnsafeRow format.
In addition, the use of Kryo as the default serializer can introduce performance problems because the creation of new KryoSerializer instances is expensive and we haven't performed instance-reuse optimizations in several code paths (including DirectTaskResult deserialization).
Given all of this, I propose to revert back to using JavaSerializer as the default serializer in the Thrift Server.
/cc liancheng
Author: Josh Rosen <joshrosen@databricks.com>
Closes#14906 from JoshRosen/disable-kryo-in-thriftserver.
## What changes were proposed in this pull request?
This patch makes RBackend connection timeout configurable by user.
## How was this patch tested?
N/A
Author: Hossein <hossein@databricks.com>
Closes#15471 from falaki/SPARK-17919.
## What changes were proposed in this pull request?
Currently users can kill stages via the web ui but not jobs directly (jobs are killed if one of their stages is). I've added the ability to kill jobs via the web ui. This code change is based on #4823 by lianhuiwang and updated to work with the latest code matching how stages are currently killed. In general I've copied the kill stage code warning and note comments and all. I also updated applicable tests and documentation.
## How was this patch tested?
Manually tested and dev/run-tests
Author: Alex Bozarth <ajbozart@us.ibm.com>
Author: Lianhui Wang <lianhuiwang09@gmail.com>
Closes#15441 from ajbozarth/spark4411.
## What changes were proposed in this pull request?
This PR adds support for executor log compression.
## How was this patch tested?
Unit tests
cc: yhuai tdas mengxr
Author: Yu Peng <loneknightpy@gmail.com>
Closes#15285 from loneknightpy/compress-executor-log.
This reverts commit ed14633414.
The patch merged had obvious quality and documentation issue. The idea is useful, and we should work towards improving its quality and merging it in again.
## What changes were proposed in this pull request?
Restructure the code and implement two new task assigner.
PackedAssigner: try to allocate tasks to the executors with least available cores, so that spark can release reserved executors when dynamic allocation is enabled.
BalancedAssigner: try to allocate tasks to the executors with more available cores in order to balance the workload across all executors.
By default, the original round robin assigner is used.
We test a pipeline, and new PackedAssigner save around 45% regarding the reserved cpu and memory with dynamic allocation enabled.
## How was this patch tested?
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
Both unit test in TaskSchedulerImplSuite and manual tests in production pipeline.
Author: Zhan Zhang <zhanzhang@fb.com>
Closes#15218 from zhzhan/packed-scheduler.
## What changes were proposed in this pull request?
This is a step along the way to SPARK-8425.
To enable incremental review, the first step proposed here is to expand the blacklisting within tasksets. In particular, this will enable blacklisting for
* (task, executor) pairs (this already exists via an undocumented config)
* (task, node)
* (taskset, executor)
* (taskset, node)
Adding (task, node) is critical to making spark fault-tolerant of one-bad disk in a cluster, without requiring careful tuning of "spark.task.maxFailures". The other additions are also important to avoid many misleading task failures and long scheduling delays when there is one bad node on a large cluster.
Note that some of the code changes here aren't really required for just this -- they put pieces in place for SPARK-8425 even though they are not used yet (eg. the `BlacklistTracker` helper is a little out of place, `TaskSetBlacklist` holds onto a little more info than it needs to for just this change, and `ExecutorFailuresInTaskSet` is more complex than it needs to be).
## How was this patch tested?
Added unit tests, run tests via jenkins.
Author: Imran Rashid <irashid@cloudera.com>
Author: mwws <wei.mao@intel.com>
Closes#15249 from squito/taskset_blacklist_only.
The goal of this feature is to allow the Spark driver to run in an
isolated environment, such as a docker container, and be able to use
the host's port forwarding mechanism to be able to accept connections
from the outside world.
The change is restricted to the driver: there is no support for achieving
the same thing on executors (or the YARN AM for that matter). Those still
need full access to the outside world so that, for example, connections
can be made to an executor's block manager.
The core of the change is simple: add a new configuration that tells what's
the address the driver should bind to, which can be different than the address
it advertises to executors (spark.driver.host). Everything else is plumbing
the new configuration where it's needed.
To use the feature, the host starting the container needs to set up the
driver's port range to fall into a range that is being forwarded; this
required the block manager port to need a special configuration just for
the driver, which falls back to the existing spark.blockManager.port when
not set. This way, users can modify the driver settings without affecting
the executors; it would theoretically be nice to also have different
retry counts for driver and executors, but given that docker (at least)
allows forwarding port ranges, we can probably live without that for now.
Because of the nature of the feature it's kinda hard to add unit tests;
I just added a simple one to make sure the configuration works.
This was tested with a docker image running spark-shell with the following
command:
docker blah blah blah \
-p 38000-38100:38000-38100 \
[image] \
spark-shell \
--num-executors 3 \
--conf spark.shuffle.service.enabled=false \
--conf spark.dynamicAllocation.enabled=false \
--conf spark.driver.host=[host's address] \
--conf spark.driver.port=38000 \
--conf spark.driver.blockManager.port=38020 \
--conf spark.ui.port=38040
Running on YARN; verified the driver works, executors start up and listen
on ephemeral ports (instead of using the driver's config), and that caching
and shuffling (without the shuffle service) works. Clicked through the UI
to make sure all pages (including executor thread dumps) worked. Also tested
apps without docker, and ran unit tests.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#15120 from vanzin/SPARK-4563.
## What changes were proposed in this pull request?
The relation between spark.network.timeout and spark.executor.heartbeatInterval should be mentioned in the document.
… network timeout]
Author: Jagadeesan <as2@us.ibm.com>
Closes#15042 from jagadeesanas2/SPARK-17449.
## What changes were proposed in this pull request?
This pull request adds the functionality to enable accessing worker and application UI through master UI itself. Thus helps in accessing SparkUI when running spark cluster in closed networks e.g. Kubernetes. Cluster admin needs to expose only spark master UI and rest of the UIs can be in the private network, master UI will reverse proxy the connection request to corresponding resource. It adds the path for workers/application UIs as
WorkerUI: <http/https>://master-publicIP:<port>/target/workerID/
ApplicationUI: <http/https>://master-publicIP:<port>/target/appID/
This makes it easy for users to easily protect the Spark master cluster access by putting some reverse proxy e.g. https://github.com/bitly/oauth2_proxy
## How was this patch tested?
The functionality has been tested manually and there is a unit test too for testing access to worker UI with reverse proxy address.
pwendell bomeng BryanCutler can you please review it, thanks.
Author: Gurvinder Singh <gurvinder.singh@uninett.no>
Closes#13950 from gurvindersingh/rproxy.
## What changes were proposed in this pull request?
Allow user to set sparkr shell command through --conf spark.r.shell.command
## How was this patch tested?
Unit test is added and also verify it manually through
```
bin/sparkr --master yarn-client --conf spark.r.shell.command=/usr/local/bin/R
```
Author: Jeff Zhang <zjffdu@apache.org>
Closes#14744 from zjffdu/SPARK-17178.
This patch is using Apache Commons Crypto library to enable shuffle encryption support.
Author: Ferdinand Xu <cheng.a.xu@intel.com>
Author: kellyzly <kellyzly@126.com>
Closes#8880 from winningsix/SPARK-10771.
## What changes were proposed in this pull request?
Based on #12990 by tankkyo
Since the History Server currently loads all application's data it can OOM if too many applications have a significant task count. `spark.ui.trimTasks` (default: false) can be set to true to trim tasks by `spark.ui.retainedTasks` (default: 10000)
(This is a "quick fix" to help those running into the problem until a update of how the history server loads app data can be done)
## How was this patch tested?
Manual testing and dev/run-tests
Author: Alex Bozarth <ajbozart@us.ibm.com>
Closes#14673 from ajbozarth/spark15083.
## What changes were proposed in this pull request?
`spark.ssl.enabled`=true, but failing to set `spark.ssl.protocol` will fail and throw meaningless exception. `spark.ssl.protocol` is required when `spark.ssl.enabled`.
Improvement: require `spark.ssl.protocol` when initializing SSLContext, otherwise throws an exception to indicate that.
Remove the OrElse("default").
Document this requirement in configure.md
## How was this patch tested?
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
Manual tests:
Build document and check document
Configure `spark.ssl.enabled` only, it throws exception below:
6/08/16 16:04:37 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(mwang); groups with view permissions: Set(); users with modify permissions: Set(mwang); groups with modify permissions: Set()
Exception in thread "main" java.lang.IllegalArgumentException: requirement failed: spark.ssl.protocol is required when enabling SSL connections.
at scala.Predef$.require(Predef.scala:224)
at org.apache.spark.SecurityManager.<init>(SecurityManager.scala:285)
at org.apache.spark.deploy.master.Master$.startRpcEnvAndEndpoint(Master.scala:1026)
at org.apache.spark.deploy.master.Master$.main(Master.scala:1011)
at org.apache.spark.deploy.master.Master.main(Master.scala)
Configure `spark.ssl.protocol` and `spark.ssl.protocol`
It works fine.
Author: wm624@hotmail.com <wm624@hotmail.com>
Closes#14674 from wangmiao1981/ssl.
## What changes were proposed in this pull request?
The configuration doc lost the config option `spark.ui.enabled` (default value is `true`)
I think this option is important because many cases we would like to turn it off.
so I add it.
## How was this patch tested?
N/A
Author: WeichenXu <WeichenXu123@outlook.com>
Closes#14604 from WeichenXu123/add_doc_param_spark_ui_enabled.
Before this PR, user have to export environment variable to specify the python of driver & executor which is not so convenient for users. This PR is trying to allow user to specify python through configuration "--pyspark-driver-python" & "--pyspark-executor-python"
Manually test in local & yarn mode for pyspark-shell and pyspark batch mode.
Author: Jeff Zhang <zjffdu@apache.org>
Closes#13146 from zjffdu/SPARK-13081.
Docs adjustment to:
- link to other relevant section of docs
- correct statement about the only value when actually other values are supported
Author: Andrew Ash <andrew@andrewash.com>
Closes#14581 from ash211/patch-10.
## What changes were proposed in this pull request?
Shuffle fetch on large intermediate dataset is slow because the shuffle service open/close the index file for each shuffle fetch. This change introduces a cache for the index information so that we can avoid accessing the index files for each block fetch
## How was this patch tested?
Tested by running a job on the cluster and the shuffle read time was reduced by 50%.
Author: Sital Kedia <skedia@fb.com>
Closes#12944 from sitalkedia/shuffle_service.
## What changes were proposed in this pull request?
Minor doc fix regarding the spark.speculation.quantile configuration parameter. It incorrectly states it should be a percentage, when it should be a fraction.
## How was this patch tested?
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
(If this patch involves UI changes, please attach a screenshot; otherwise, remove this)
I tried building the documentation but got some unidoc errors. I also got them when building off origin/master, so I don't think I caused that problem. I did run the web app and saw the changes reflected as expected.
Author: Nicholas Brown <nbrown@adroitdigital.com>
Closes#14352 from nwbvt/master.
Clarify documentation on spark.task.maxFailures
No tests run as its documentation
Author: Tom Graves <tgraves@yahoo-inc.com>
Closes#14287 from tgravescs/SPARK-16650.
## What changes were proposed in this pull request?
I search the whole documents directory using SQLContext, and update the following places:
- docs/configuration.md, sparkR code snippets.
- docs/streaming-programming-guide.md, several example code.
## How was this patch tested?
N/A
Author: WeichenXu <WeichenXu123@outlook.com>
Closes#14025 from WeichenXu123/WIP_SQLContext_update.
## What changes were proposed in this pull request?
This changes the behavior of --num-executors and spark.executor.instances when using dynamic allocation. Instead of turning dynamic allocation off, it uses the value for the initial number of executors.
This changes was discussed on [SPARK-13723](https://issues.apache.org/jira/browse/SPARK-13723). I highly recommend using it while we can change the behavior for 2.0.0. In practice, the 1.x behavior causes unexpected behavior for users (it is not clear that it disables dynamic allocation) and wastes cluster resources because users rarely notice the log message.
## How was this patch tested?
This patch updates tests and adds a test for Utils.getDynamicAllocationInitialExecutors.
Author: Ryan Blue <blue@apache.org>
Closes#13338 from rdblue/SPARK-13723-num-executors-with-dynamic-allocation.
## What changes were proposed in this pull request?
Reduce `spark.memory.fraction` default to 0.6 in order to make it fit within default JVM old generation size (2/3 heap). See JIRA discussion. This means a full cache doesn't spill into the new gen. CC andrewor14
## How was this patch tested?
Jenkins tests.
Author: Sean Owen <sowen@cloudera.com>
Closes#13618 from srowen/SPARK-15796.
While there, also document spark.files and spark.jars. Text is the
same as the spark-submit help text with some minor adjustments.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#13502 from vanzin/SPARK-15760.
## What changes were proposed in this pull request?
The configuration setting `spark.executor.logs.rolling.size.maxBytes` was changed to `spark.executor.logs.rolling.maxSize` in 1.4 or so.
This commit fixes a remaining reference to the old name in the documentation.
Also the description for `spark.executor.logs.rolling.maxSize` was edited to clearly state that the unit for the size is bytes.
## How was this patch tested?
no tests
Author: Philipp Hoffmann <mail@philipphoffmann.de>
Closes#13001 from philipphoffmann/patch-3.
## What changes were proposed in this pull request?
Currently only a list of users can be specified for view and modify acls. This change enables a group of admins/devs/users to be provisioned for viewing and modifying Spark jobs.
**Changes Proposed in the fix**
Three new corresponding config entries have been added where the user can specify the groups to be given access.
```
spark.admin.acls.groups
spark.modify.acls.groups
spark.ui.view.acls.groups
```
New config entries were added because specifying the users and groups explicitly is a better and cleaner way compared to specifying them in the existing config entry using a delimiter.
A generic trait has been introduced to provide the user to group mapping which makes it pluggable to support a variety of mapping protocols - similar to the one used in hadoop. A default unix shell based implementation has been provided.
Custom user to group mapping protocol can be specified and configured by the entry ```spark.user.groups.mapping```
**How the patch was Tested**
We ran different spark jobs setting the config entries in combinations of admin, modify and ui acls. For modify acls we tried killing the job stages from the ui and using yarn commands. For view acls we tried accessing the UI tabs and the logs. Headless accounts were used to launch these jobs and different users tried to modify and view the jobs to ensure that the groups mapping applied correctly.
Additional Unit tests have been added without modifying the existing ones. These test for different ways of setting the acls through configuration and/or API and validate the expected behavior.
Author: Dhruve Ashar <dhruveashar@gmail.com>
Closes#12760 from dhruve/impr/SPARK-4224.
## What changes were proposed in this pull request?
The sort shuffle manager has been the default since Spark 1.2. It is time to remove the old hash shuffle manager.
## How was this patch tested?
Removed some tests related to the old manager.
Author: Reynold Xin <rxin@databricks.com>
Closes#12423 from rxin/SPARK-14667.
## What changes were proposed in this pull request?
The configuration docs are updated to reflect the changes introduced with [SPARK-12384](https://issues.apache.org/jira/browse/SPARK-12384). This allows the user to specify initial heap memory settings through the extraJavaOptions for executor, driver and am.
## How was this patch tested?
The changes are tested in [SPARK-12384](https://issues.apache.org/jira/browse/SPARK-12384). This is just documenting the changes made.
Author: Dhruve Ashar <dhruveashar@gmail.com>
Closes#12333 from dhruve/doc/SPARK-14572.
The description of "spark.memory.offHeap.size" in the current document does not clearly state that memory is counted with bytes....
This PR contains a small fix for this tiny issue
document fix
Author: CodingCat <zhunansjtu@gmail.com>
Closes#11561 from CodingCat/master.
## What changes were proposed in this pull request?
We provide a very limited set of cluster management script in Spark for Tachyon, although Tachyon itself provides a much better version of it. Given now Spark users can simply use Tachyon as a normal file system and does not require extensive configurations, we can remove this management capabilities to simplify Spark bash scripts.
Note that this also reduces coupling between a 3rd party external system and Spark's release scripts, and would eliminate possibility for failures such as Tachyon being renamed or the tar balls being relocated.
## How was this patch tested?
N/A
Author: Reynold Xin <rxin@databricks.com>
Closes#11400 from rxin/release-script.
## What changes were proposed in this pull request?
This PR fixes some typos in the following documentation files.
* `NOTICE`, `configuration.md`, and `hardware-provisioning.md`.
## How was the this patch tested?
manual tests
Author: Dongjoon Hyun <dongjoonapache.org>
Author: Dongjoon Hyun <dongjoon@apache.org>
Closes#11289 from dongjoon-hyun/minor_fix_typos_notice_and_confdoc.
https://issues.apache.org/jira/browse/SPARK-11627
Spark Streaming backpressure mechanism has no initial input rate limit, it might cause OOM exception.
In the firest batch task ,receivers receive data at the maximum speed they can reach,it might exhaust executors memory resources. Add a initial input rate limit value can make sure the Streaming job execute success in the first batch,then the backpressure mechanism can adjust receiving rate adaptively.
Author: junhao <junhao@mogujie.com>
Closes#9593 from junhaoMg/junhao-dev.
This JIRA is related to
https://github.com/apache/spark/pull/5852
Had to do some minor rework and test to make sure it
works with current version of spark.
Author: Sanket <schintap@untilservice-lm>
Closes#10838 from redsanket/limit-outbound-connections.
Remove spark.closure.serializer option and use JavaSerializer always
CC andrewor14 rxin I see there's a discussion in the JIRA but just thought I'd offer this for a look at what the change would be.
Author: Sean Owen <sowen@cloudera.com>
Closes#11150 from srowen/SPARK-12414.
This is the next iteration of tnachen's previous PR: https://github.com/apache/spark/pull/4027
In that PR, we resolved with andrewor14 and pwendell to implement the Mesos scheduler's support of `spark.executor.cores` to be consistent with YARN and Standalone. This PR implements that resolution.
This PR implements two high-level features. These two features are co-dependent, so they're implemented both here:
- Mesos support for spark.executor.cores
- Multiple executors per slave
We at Mesosphere have been working with Typesafe on a Spark/Mesos integration test suite: https://github.com/typesafehub/mesos-spark-integration-tests, which passes for this PR.
The contribution is my original work and I license the work to the project under the project's open source license.
Author: Michael Gummelt <mgummelt@mesosphere.io>
Closes#10993 from mgummelt/executor_sizing.
Fix zookeeper dir configuration used in cluster mode, and also add documentation around these settings.
Author: Timothy Chen <tnachen@gmail.com>
Closes#10057 from tnachen/fix_mesos_dir.
JIRA 1680 added a property called spark.yarn.appMasterEnv. This PR draws users' attention to this special case by adding an explanation in configuration.html#environment-variables
Author: Andrew <weiner.andrew.j@gmail.com>
Closes#10869 from weineran/branch-yarn-docs.
- Remove Akka dependency from core. Note: the streaming-akka project still uses Akka.
- Remove HttpFileServer
- Remove Akka configs from SparkConf and SSLOptions
- Rename `spark.akka.frameSize` to `spark.rpc.message.maxSize`. I think it's still worth to keep this config because using `DirectTaskResult` or `IndirectTaskResult` depends on it.
- Update comments and docs
Author: Shixiong Zhu <shixiong@databricks.com>
Closes#10854 from zsxwing/remove-akka.
Several Spark properties equivalent to Spark submit command line options are missing.
Author: felixcheung <felixcheung_m@hotmail.com>
Closes#10491 from felixcheung/sparksubmitdoc.
modify 'spark.memory.offHeap.enabled' default value to false
Author: zzcclp <xm_zzc@sina.com>
Closes#10633 from zzcclp/fix_spark.memory.offHeap.enabled_default_value.
This PR removes `spark.cleaner.ttl` and the associated TTL-based metadata cleaning code.
Now that we have the `ContextCleaner` and a timer to trigger periodic GCs, I don't think that `spark.cleaner.ttl` is necessary anymore. The TTL-based cleaning isn't enabled by default, isn't included in our end-to-end tests, and has been a source of user confusion when it is misconfigured. If the TTL is set too low, data which is still being used may be evicted / deleted, leading to hard to diagnose bugs.
For all of these reasons, I think that we should remove this functionality in Spark 2.0. Additional benefits of doing this include marginally reduced memory usage, since we no longer need to store timetsamps in hashmaps, and a handful fewer threads.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#10534 from JoshRosen/remove-ttl-based-cleaning.
We switched to TorrentBroadcast in Spark 1.1, and HttpBroadcast has been undocumented since then. It's time to remove it in Spark 2.0.
Author: Reynold Xin <rxin@databricks.com>
Closes#10531 from rxin/SPARK-12588.
According the benchmark [1], LZ4-java could be 80% (or 30%) faster than Snappy.
After changing the compressor to LZ4, I saw 20% improvement on end-to-end time for a TPCDS query (Q4).
[1] https://github.com/ning/jvm-compressor-benchmark/wiki
cc rxin
Author: Davies Liu <davies@databricks.com>
Closes#10342 from davies/lz4.
The current default storage level of Python persist API is MEMORY_ONLY_SER. This is different from the default level MEMORY_ONLY in the official document and RDD APIs.
davies Is this inconsistency intentional? Thanks!
Updates: Since the data is always serialized on the Python side, the storage levels of JAVA-specific deserialization are not removed, such as MEMORY_ONLY.
Updates: Based on the reviewers' feedback. In Python, stored objects will always be serialized with the [Pickle](https://docs.python.org/2/library/pickle.html) library, so it does not matter whether you choose a serialized level. The available storage levels in Python include `MEMORY_ONLY`, `MEMORY_ONLY_2`, `MEMORY_AND_DISK`, `MEMORY_AND_DISK_2`, `DISK_ONLY`, `DISK_ONLY_2` and `OFF_HEAP`.
Author: gatorsmile <gatorsmile@gmail.com>
Closes#10092 from gatorsmile/persistStorageLevel.
This patch adds documentation for Spark configurations that affect off-heap memory and makes some naming and validation improvements for those configs.
- Change `spark.memory.offHeapSize` to `spark.memory.offHeap.size`. This is fine because this configuration has not shipped in any Spark release yet (it's new in Spark 1.6).
- Deprecated `spark.unsafe.offHeap` in favor of a new `spark.memory.offHeap.enabled` configuration. The motivation behind this change is to gather all memory-related configurations under the same prefix.
- Add a check which prevents users from setting `spark.memory.offHeap.enabled=true` when `spark.memory.offHeap.size == 0`. After SPARK-11389 (#9344), which was committed in Spark 1.6, Spark enforces a hard limit on the amount of off-heap memory that it will allocate to tasks. As a result, enabling off-heap execution memory without setting `spark.memory.offHeap.size` will lead to immediate OOMs. The new configuration validation makes this scenario easier to diagnose, helping to avoid user confusion.
- Document these configurations on the configuration page.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#10237 from JoshRosen/SPARK-12251.
This avoids bringing up yet another HTTP server on the driver, and
instead reuses the file server already managed by the driver's
RpcEnv. As a bonus, the repl now inherits the security features of
the network library.
There's also a small change to create the directory for storing classes
under the root temp dir for the application (instead of directly
under java.io.tmpdir).
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#9923 from vanzin/SPARK-11563.
The existing `spark.memory.fraction` (default 0.75) gives the system 25% of the space to work with. For small heaps, this is not enough: e.g. default 1GB leaves only 250MB system memory. This is especially a problem in local mode, where the driver and executor are crammed in the same JVM. Members of the community have reported driver OOM's in such cases.
**New proposal.** We now reserve 300MB before taking the 75%. For 1GB JVMs, this leaves `(1024 - 300) * 0.75 = 543MB` for execution and storage. This is proposal (1) listed in the [JIRA](https://issues.apache.org/jira/browse/SPARK-12081).
Author: Andrew Or <andrew@databricks.com>
Closes#10081 from andrewor14/unified-memory-small-heaps.
This change abstracts the code that serves jars / files to executors so that
each RpcEnv can have its own implementation; the akka version uses the existing
HTTP-based file serving mechanism, while the netty versions uses the new
stream support added to the network lib, which makes file transfers benefit
from the easier security configuration of the network library, and should also
reduce overhead overall.
The change includes a small fix to TransportChannelHandler so that it propagates
user events to downstream handlers.
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#9530 from vanzin/SPARK-11140.
`<\code>` end tag missing backslash in
docs/configuration.md{L308-L339}
ref #8795
Author: Kai Jiang <jiangkai@gmail.com>
Closes#9715 from vectorijk/minor-typo-docs.
Remove Hadoop third party distro page, and move Hadoop cluster config info to configuration page
CC pwendell
Author: Sean Owen <sowen@cloudera.com>
Closes#9298 from srowen/SPARK-11305.
There's a lot of duplication between SortShuffleManager and UnsafeShuffleManager. Given that these now provide the same set of functionality, now that UnsafeShuffleManager supports large records, I think that we should replace SortShuffleManager's serialized shuffle implementation with UnsafeShuffleManager's and should merge the two managers together.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#8829 from JoshRosen/consolidate-sort-shuffle-implementations.
Add documentation for configuration:
- spark.sql.ui.retainedExecutions
- spark.streaming.ui.retainedBatches
Author: Nick Pritchard <nicholas.pritchard@falkonry.com>
Closes#9052 from pnpritchard/SPARK-11039.
This patch unifies the memory management of the storage and execution regions such that either side can borrow memory from each other. When memory pressure arises, storage will be evicted in favor of execution. To avoid regressions in cases where storage is crucial, we dynamically allocate a fraction of space for storage that execution cannot evict. Several configurations are introduced:
- **spark.memory.fraction (default 0.75)**: fraction of the heap space used for execution and storage. The lower this is, the more frequently spills and cached data eviction occur. The purpose of this config is to set aside memory for internal metadata, user data structures, and imprecise size estimation in the case of sparse, unusually large records.
- **spark.memory.storageFraction (default 0.5)**: size of the storage region within the space set aside by `spark.memory.fraction`. Cached data may only be evicted if total storage exceeds this region.
- **spark.memory.useLegacyMode (default false)**: whether to use the memory management that existed in Spark 1.5 and before. This is mainly for backward compatibility.
For a detailed description of the design, see [SPARK-10000](https://issues.apache.org/jira/browse/SPARK-10000). This patch builds on top of the `MemoryManager` interface introduced in #9000.
Author: Andrew Or <andrew@databricks.com>
Closes#9084 from andrewor14/unified-memory-manager.
1.4 docs noted that the units were MB - i have assumed this is still the case
Author: admackin <admackin@users.noreply.github.com>
Closes#9025 from admackin/master.
It does not make much sense to set `spark.shuffle.spill` or `spark.sql.planner.externalSort` to false: I believe that these configurations were initially added as "escape hatches" to guard against bugs in the external operators, but these operators are now mature and well-tested. In addition, these configurations are not handled in a consistent way anymore: SQL's Tungsten codepath ignores these configurations and will continue to use spilling operators. Similarly, Spark Core's `tungsten-sort` shuffle manager does not respect `spark.shuffle.spill=false`.
This pull request removes these configurations, adds warnings at the appropriate places, and deletes a large amount of code which was only used in code paths that did not support spilling.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#8831 from JoshRosen/remove-ability-to-disable-spilling.
spark.scheduler.minRegisteredResourcesRatio configuration parameter works for YARN mode but not for Mesos Coarse grained mode.
If the parameter specified default value of 0 will be set for spark.scheduler.minRegisteredResourcesRatio in base class and this method will always return true.
There are no existing test for YARN mode too. Hence not added test for the same.
Author: Akash Mishra <akash.mishra20@gmail.com>
Closes#8672 from SleepyThread/master.
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#8656 from tdas/SPARK-10492 and squashes the following commits:
986cdd6 [Tathagata Das] Added information on backpressure
We introduced the Netty network module for shuffle in Spark 1.2, and has turned it on by default for 3 releases. The old ConnectionManager is difficult to maintain. If we merge the patch now, by the time it is released, it would be 1 yr for which ConnectionManager is off by default. It's time to remove it.
Author: Reynold Xin <rxin@databricks.com>
Closes#8161 from rxin/SPARK-9767.
SPARK-4223.
Currently we support setting view and modify acls but you have to specify a list of users. It would be nice to support * meaning all users have access.
Manual tests to verify that: "*" works for any user in:
a. Spark ui: view and kill stage. Done.
b. Spark history server. Done.
c. Yarn application killing. Done.
Author: zhuol <zhuol@yahoo-inc.com>
Closes#8398 from zhuoliu/4223.
https://issues.apache.org/jira/browse/SPARK-10315
this parameter is not used any longer and there is some mistake in the current document , should be 'akka.remote.watch-failure-detector.threshold'
Author: CodingCat <zhunansjtu@gmail.com>
Closes#8483 from CodingCat/SPARK_10315.
Deprecate NIO ConnectionManager in Spark 1.5.0, before removing it in Spark 1.6.0.
Author: Reynold Xin <rxin@databricks.com>
Closes#8162 from rxin/SPARK-9934.
Document spark.shuffle.service.{enabled,port}
CC sryza tgravescs
This is pretty minimal; is there more to say here about the service?
Author: Sean Owen <sowen@cloudera.com>
Closes#7991 from srowen/SPARK-9641 and squashes the following commits:
3bb946e [Sean Owen] Add link to docs for setup and config of external shuffle service
2302e01 [Sean Owen] Document spark.shuffle.service.{enabled,port}
Author: Marcelo Vanzin <vanzin@cloudera.com>
Closes#7651 from vanzin/SPARK-9327 and squashes the following commits:
2923e23 [Marcelo Vanzin] [SPARK-9327] [docs] Fix documentation about classpath config options.
Spark has an option called spark.localExecution.enabled; according to the docs:
> Enables Spark to run certain jobs, such as first() or take() on the driver, without sending tasks to the cluster. This can make certain jobs execute very quickly, but may require shipping a whole partition of data to the driver.
This feature ends up adding quite a bit of complexity to DAGScheduler, especially in the runLocallyWithinThread method, but as far as I know nobody uses this feature (I searched the mailing list and haven't seen any recent mentions of the configuration nor stacktraces including the runLocally method). As a step towards scheduler complexity reduction, I propose that we remove this feature and all code related to it for Spark 1.5.
This pull request simply brings #7484 up to date.
Author: Josh Rosen <joshrosen@databricks.com>
Author: Reynold Xin <rxin@databricks.com>
Closes#7585 from rxin/remove-local-exec and squashes the following commits:
84bd10e [Reynold Xin] Python fix.
1d9739a [Reynold Xin] Merge pull request #7484 from JoshRosen/remove-localexecution
eec39fa [Josh Rosen] Remove allowLocal(); deprecate user-facing uses of it.
b0835dc [Josh Rosen] Remove local execution code in DAGScheduler
8975d96 [Josh Rosen] Remove local execution tests.
ffa8c9b [Josh Rosen] Remove documentation for configuration
There are a few memory limits that people hit often and that we could
make higher, especially now that memory sizes have grown.
- spark.akka.frameSize: This defaults at 10 but is often hit for map
output statuses in large shuffles. This memory is not fully allocated
up-front, so we can just make this larger and still not affect jobs
that never sent a status that large. We increase it to 128.
- spark.executor.memory: Defaults at 512m, which is really small. We
increase it to 1g.
Author: Matei Zaharia <matei@databricks.com>
Closes#7586 from mateiz/configs and squashes the following commits:
ce0038a [Matei Zaharia] [SPARK-9244] Increase some memory defaults
The meaning of spark.kryoserializer.buffer should be "Initial size of Kryo's serialization buffer. Note that there will be one buffer per core on each worker. This buffer will grow up to spark.kryoserializer.buffer.max if needed.".
The spark.kryoserializer.buffer.max.mb is out-of-date in spark 1.4.
Author: zhaishidan <zhaishidan@haizhi.com>
Closes#7393 from stanzhai/master and squashes the following commits:
69729ef [zhaishidan] fix document error about spark.kryoserializer.buffer.max.mb
pwendell and I discussed this a little more offline and concluded that it would be good to keep it more conservative. Losing cached blocks may be very expensive and we should only allow it if the user knows what he/she is doing.
FYI harishreedharan sryza.
Author: Andrew Or <andrew@databricks.com>
Closes#7329 from andrewor14/da-cached-timeout and squashes the following commits:
cef0b4e [Andrew Or] Change timeout to infinity
A couple descriptions were not inside `<td></td>` and were being displayed immediately under the section title instead of in their row.
Author: Jonathan Alter <jonalter@users.noreply.github.com>
Closes#7292 from jonalter/docs-config and squashes the following commits:
5ce1570 [Jonathan Alter] [DOCS] Format wrong for some config descriptions
I've updated default values in comments, documentation, and in the command line builder to be 1g based on comments in the JIRA. I've also updated most usages to point at a single variable defined in the Utils.scala and JavaUtils.java files. This wasn't possible in all cases (R, shell scripts etc.) but usage in most code is now pointing at the same place.
Please let me know if I've missed anything.
Will the spark-shell use the value within the command line builder during instantiation?
Author: Ilya Ganelin <ilya.ganelin@capitalone.com>
Closes#7132 from ilganeli/SPARK-3071 and squashes the following commits:
4074164 [Ilya Ganelin] String fix
271610b [Ilya Ganelin] Merge branch 'SPARK-3071' of github.com:ilganeli/spark into SPARK-3071
273b6e9 [Ilya Ganelin] Test fix
fd67721 [Ilya Ganelin] Update JavaUtils.java
26cc177 [Ilya Ganelin] test fix
e5db35d [Ilya Ganelin] Fixed test failure
39732a1 [Ilya Ganelin] merge fix
a6f7deb [Ilya Ganelin] Created default value for DRIVER MEM in Utils that's now used in almost all locations instead of setting manually in each
09ad698 [Ilya Ganelin] Update SubmitRestProtocolSuite.scala
19b6f25 [Ilya Ganelin] Missed one doc update
2698a3d [Ilya Ganelin] Updated default value for driver memory
[SQL][DOC] I found it a bit confusing when I came across it for the first time in the docs
Author: Radek Ostrowski <dest.hawaii@gmail.com>
Author: radek <radek@radeks-MacBook-Pro-2.local>
Closes#6332 from radek1st/master and squashes the following commits:
dae3347 [Radek Ostrowski] fixed typo
c76bb3a [radek] improved a comment
Read number of threads for RBackend from configuration.
[SPARK-8282] #comment Linking with JIRA
Author: Hossein <hossein@databricks.com>
Closes#6730 from falaki/SPARK-8282 and squashes the following commits:
33b3d98 [Hossein] Documented new config parameter
70f2a9c [Hossein] Fixing import
ec44225 [Hossein] Read number of threads for RBackend from configuration
this is a follow up of #3621
/cc liancheng pwendell
Author: Daoyuan Wang <daoyuan.wang@intel.com>
Closes#6639 from adrian-wang/kryodoc and squashes the following commits:
3c4b1cf [Daoyuan Wang] [DOC] kryo default setting in SQL Thrift server
…moved if dynamic allocation is enabled.
This is a work in progress. This patch ensures that an executor that has cached RDD blocks are not removed,
but makes no attempt to find another executor to remove. This is meant to get some feedback on the current
approach, and if it makes sense then I will look at choosing another executor to remove. No testing has been done either.
Author: Hari Shreedharan <hshreedharan@apache.org>
Closes#6508 from harishreedharan/dymanic-caching and squashes the following commits:
dddf1eb [Hari Shreedharan] Minor configuration description update.
10130e2 [Hari Shreedharan] Fix compile issue.
5417b53 [Hari Shreedharan] Add documentation for new config. Remove block from cachedBlocks when it is dropped.
875916a [Hari Shreedharan] Make some code more readable.
39940ca [Hari Shreedharan] Handle the case where the executor has not yet registered.
90ad711 [Hari Shreedharan] Remove unused imports and unused methods.
063985c [Hari Shreedharan] Send correct message instead of recursively calling same method.
ec2fd7e [Hari Shreedharan] Add file missed in last commit
5d10fad [Hari Shreedharan] Update cached blocks status using local info, rather than doing an RPC.
193af4c [Hari Shreedharan] WIP. Use local state rather than via RPC.
ae932ff [Hari Shreedharan] Fix config param name.
272969d [Hari Shreedharan] Fix seconds to millis bug.
5a1993f [Hari Shreedharan] Add timeout for cache executors. Ignore broadcast blocks while checking if there are cached blocks.
57fefc2 [Hari Shreedharan] [SPARK-7955][Core] Ensure executors with cached RDD blocks are not removed if dynamic allocation is enabled.
The first line had only two dashes (--) instead of three(---). Because of this missing dash(-), 'jekyll build' command was not converting configuration.md to _site/configuration.html
Author: Taka Shinagawa <taka.epsilon@gmail.com>
Closes#6513 from mrt/docfix3 and squashes the following commits:
c470e2c [Taka Shinagawa] Added a missing dash(-) preventing jekyll from converting configuration.md to html format
The default add time of 5s is still too slow for small jobs. Also, the current default remove time of 10 minutes seem rather high. This patch lowers both and rephrases a few log messages.
Author: Andrew Or <andrew@databricks.com>
Closes#6301 from andrewor14/da-minor and squashes the following commits:
6d614a6 [Andrew Or] Lower log level
2811492 [Andrew Or] Log information when requests are canceled
5fcd3eb [Andrew Or] Fix tests
3320710 [Andrew Or] Lower timeouts + rephrase a few log messages
The default value will be changed to `1000` in #5533. So here I just used `1000`.
Author: zsxwing <zsxwing@gmail.com>
Closes#5899 from zsxwing/SPARK-7351 and squashes the following commits:
e1ec515 [zsxwing] [SPARK-7351][Streaming][Docs] Add spark.streaming.ui.retainedBatches to docs
Added documentation for spark.streaming.kafka.maxRetries
Author: BenFradet <benjamin.fradet@gmail.com>
Closes#5808 from BenFradet/master and squashes the following commits:
cc72e7a [BenFradet] updated doc for spark.streaming.kafka.maxRetries to explain the default value
18f823e [BenFradet] Added "consecutive" to the spark.streaming.kafka.maxRetries doc
597fdeb [BenFradet] Mention that spark.streaming.kafka.maxRetries only applies to the direct kafka api
0efad39 [BenFradet] Added documentation for spark.streaming.kafka.maxRetries
This is the classes for creating off-heap block storage API. It also includes the migration for Tachyon. The diff seems to be big, but it mainly just rename tachyon to offheap. New implementation for hdfs will be submit for review in spark-6112.
Author: Zhan Zhang <zhazhan@gmail.com>
Closes#5430 from zhzhan/SPARK-6479 and squashes the following commits:
60acd84 [Zhan Zhang] minor change to kickoff the test
12f54c9 [Zhan Zhang] solve merge conflicts
a54132c [Zhan Zhang] solve review comments
ffb8e00 [Zhan Zhang] rebase to sparkcontext change
6e121e0 [Zhan Zhang] resolve review comments and restructure blockmanasger code
a7aed6c [Zhan Zhang] add Tachyon migration code
186de31 [Zhan Zhang] initial commit for off-heap block storage api
I've added an interface to JavaUtils to do byte conversion and added hooks within Utils.scala to handle conversion within Spark code (like for time strings). I've added matching tests for size conversion, and then updated all deprecated configs and documentation as per SPARK-5933.
Author: Ilya Ganelin <ilya.ganelin@capitalone.com>
Closes#5574 from ilganeli/SPARK-5932 and squashes the following commits:
11f6999 [Ilya Ganelin] Nit fixes
49a8720 [Ilya Ganelin] Whitespace fix
2ab886b [Ilya Ganelin] Scala style
fc85733 [Ilya Ganelin] Got rid of floating point math
852a407 [Ilya Ganelin] [SPARK-5932] Added much improved overflow handling. Can now handle sizes up to Long.MAX_VALUE Petabytes instead of being capped at Long.MAX_VALUE Bytes
9ee779c [Ilya Ganelin] Simplified fraction matches
22413b1 [Ilya Ganelin] Made MAX private
3dfae96 [Ilya Ganelin] Fixed some nits. Added automatic conversion of old paramter for kryoserializer.mb to new values.
e428049 [Ilya Ganelin] resolving merge conflict
8b43748 [Ilya Ganelin] Fixed error in pattern matching for doubles
84a2581 [Ilya Ganelin] Added smoother handling of fractional values for size parameters. This now throws an exception and added a warning for old spark.kryoserializer.buffer
d3d09b6 [Ilya Ganelin] [SPARK-5932] Fixing error in KryoSerializer
fe286b4 [Ilya Ganelin] Resolved merge conflict
c7803cd [Ilya Ganelin] Empty lines
54b78b4 [Ilya Ganelin] Simplified byteUnit class
69e2f20 [Ilya Ganelin] Updates to code
f32bc01 [Ilya Ganelin] [SPARK-5932] Fixed error in API in SparkConf.scala where Kb conversion wasn't being done properly (was Mb). Added test cases for both timeUnit and ByteUnit conversion
f15f209 [Ilya Ganelin] Fixed conversion of kryo buffer size
0f4443e [Ilya Ganelin] Merge remote-tracking branch 'upstream/master' into SPARK-5932
35a7fa7 [Ilya Ganelin] Minor formatting
928469e [Ilya Ganelin] [SPARK-5932] Converted some longs to ints
5d29f90 [Ilya Ganelin] [SPARK-5932] Finished documentation updates
7a6c847 [Ilya Ganelin] [SPARK-5932] Updated spark.shuffle.file.buffer
afc9a38 [Ilya Ganelin] [SPARK-5932] Updated spark.broadcast.blockSize and spark.storage.memoryMapThreshold
ae7e9f6 [Ilya Ganelin] [SPARK-5932] Updated spark.io.compression.snappy.block.size
2d15681 [Ilya Ganelin] [SPARK-5932] Updated spark.executor.logs.rolling.size.maxBytes
1fbd435 [Ilya Ganelin] [SPARK-5932] Updated spark.broadcast.blockSize
eba4de6 [Ilya Ganelin] [SPARK-5932] Updated spark.shuffle.file.buffer.kb
b809a78 [Ilya Ganelin] [SPARK-5932] Updated spark.kryoserializer.buffer.max
0cdff35 [Ilya Ganelin] [SPARK-5932] Updated to use bibibytes in method names. Updated spark.kryoserializer.buffer.mb and spark.reducer.maxMbInFlight
475370a [Ilya Ganelin] [SPARK-5932] Simplified ByteUnit code, switched to using longs. Updated docs to clarify that we use kibi, mebi etc instead of kilo, mega
851d691 [Ilya Ganelin] [SPARK-5932] Updated memoryStringToMb to use new interfaces
a9f4fcf [Ilya Ganelin] [SPARK-5932] Added unit tests for unit conversion
747393a [Ilya Ganelin] [SPARK-5932] Added unit tests for ByteString conversion
09ea450 [Ilya Ganelin] [SPARK-5932] Added byte string conversion to Jav utils
5390fd9 [Ilya Ganelin] Merge remote-tracking branch 'upstream/master' into SPARK-5932
db9a963 [Ilya Ganelin] Closing second spark context
1dc0444 [Ilya Ganelin] Added ref equality check
8c884fa [Ilya Ganelin] Made getOrCreate synchronized
cb0c6b7 [Ilya Ganelin] Doc updates and code cleanup
270cfe3 [Ilya Ganelin] [SPARK-6703] Documentation fixes
15e8dea [Ilya Ganelin] Updated comments and added MiMa Exclude
0e1567c [Ilya Ganelin] Got rid of unecessary option for AtomicReference
dfec4da [Ilya Ganelin] Changed activeContext to AtomicReference
733ec9f [Ilya Ganelin] Fixed some bugs in test code
8be2f83 [Ilya Ganelin] Replaced match with if
e92caf7 [Ilya Ganelin] [SPARK-6703] Added test to ensure that getOrCreate both allows creation, retrieval, and a second context if desired
a99032f [Ilya Ganelin] Spacing fix
d7a06b8 [Ilya Ganelin] Updated SparkConf class to add getOrCreate method. Started test suite implementation
Added docs for rpc configurations and also fixed two places that should have been fixed in #5595.
Author: zsxwing <zsxwing@gmail.com>
Closes#5607 from zsxwing/SPARK-6490-docs and squashes the following commits:
25a6736 [zsxwing] Increase the default timeout to 120s
6e37c30 [zsxwing] Update docs
5577540 [zsxwing] Use spark.network.timeout as the default timeout if it presents
4f07174 [zsxwing] Fix unit tests
1c2cf26 [zsxwing] Add docs for rpc configurations
resubmit of https://github.com/apache/spark/pull/636 for a totally different algorithm
https://issues.apache.org/jira/browse/SPARK-1706
In current implementation, the user has to start multiple workers in a server for starting multiple executors in a server, which introduces additional overhead due to the more JVM processes...
In this patch, I changed the scheduling logic in master to enable the user to start multiple executor processes within the same JVM process.
1. user configure spark.executor.maxCoreNumPerExecutor to suggest the maximum core he/she would like to allocate to each executor
2. Master assigns the executors to the workers with the major consideration on the memoryPerExecutor and the worker.freeMemory, and tries to allocate as many as possible cores to the executor ```min(min(memoryPerExecutor, worker.freeCore), maxLeftCoreToAssign)``` where ```maxLeftCoreToAssign = maxExecutorCanAssign * maxCoreNumPerExecutor```
---------------------------------------
Other small changes include
change memoryPerSlave in ApplicationDescription to memoryPerExecutor, as "Slave" is overrided to represent both worker and executor in the documents... (we have some discussion on this before?)
Author: CodingCat <zhunansjtu@gmail.com>
Closes#731 from CodingCat/SPARK-1706-2 and squashes the following commits:
6dee808 [CodingCat] change filter predicate
fbeb7e5 [CodingCat] address the comments
940cb42 [CodingCat] avoid unnecessary allocation
b8ca561 [CodingCat] revert a change
45967b4 [CodingCat] remove unused method
2eeff77 [CodingCat] stylistic fixes
12a1b32 [CodingCat] change the semantic of coresPerExecutor to exact core number
f035423 [CodingCat] stylistic fix
d9c1685 [CodingCat] remove unused var
f595bd6 [CodingCat] recover some unintentional changes
63b3df9 [CodingCat] change the description of the parameter in the submit script
4cf61f1 [CodingCat] improve the code and docs
ff011e2 [CodingCat] start multiple executors on the worker by rewriting startExeuctor logic
2c2bcc5 [CodingCat] fix wrong usage info
497ec2c [CodingCat] address andrew's comments
878402c [CodingCat] change the launching executor code
f64a28d [CodingCat] typo fix
387f4ec [CodingCat] bug fix
35c462c [CodingCat] address Andrew's comments
0b64fea [CodingCat] fix compilation issue
19d3da7 [CodingCat] address the comments
5b81466 [CodingCat] remove outdated comments
ec7d421 [CodingCat] test commit
e5efabb [CodingCat] more java docs and consolidate canUse function
a26096d [CodingCat] stylistic fix
a5d629a [CodingCat] java doc
b34ec0c [CodingCat] make master support multiple executors per worker
I've added new utility methods to do the conversion from times specified as e.g. 120s, 240ms, 360us to convert to a consistent internal representation. I've updated usage of these constants throughout the code to be consistent.
I believe I've captured all usages of time-based properties throughout the code. I've also updated variable names in a number of places to reflect their units for clarity and updated documentation where appropriate.
Author: Ilya Ganelin <ilya.ganelin@capitalone.com>
Author: Ilya Ganelin <ilganeli@gmail.com>
Closes#5236 from ilganeli/SPARK-5931 and squashes the following commits:
4526c81 [Ilya Ganelin] Update configuration.md
de3bff9 [Ilya Ganelin] Fixing style errors
f5fafcd [Ilya Ganelin] Doc updates
951ca2d [Ilya Ganelin] Made the most recent round of changes
bc04e05 [Ilya Ganelin] Minor fixes and doc updates
25d3f52 [Ilya Ganelin] Minor nit fixes
642a06d [Ilya Ganelin] Fixed logic for invalid suffixes and addid matching test
8927e66 [Ilya Ganelin] Fixed handling of -1
69fedcc [Ilya Ganelin] Added test for zero
dc7bd08 [Ilya Ganelin] Fixed error in exception handling
7d19cdd [Ilya Ganelin] Added fix for possible NPE
6f651a8 [Ilya Ganelin] Now using regexes to simplify code in parseTimeString. Introduces getTimeAsSec and getTimeAsMs methods in SparkConf. Updated documentation
cbd2ca6 [Ilya Ganelin] Formatting error
1a1122c [Ilya Ganelin] Formatting fixes and added m for use as minute formatter
4e48679 [Ilya Ganelin] Fixed priority order and mixed up conversions in a couple spots
d4efd26 [Ilya Ganelin] Added time conversion for yarn.scheduler.heartbeat.interval-ms
cbf41db [Ilya Ganelin] Got rid of thrown exceptions
1465390 [Ilya Ganelin] Nit
28187bf [Ilya Ganelin] Convert straight to seconds
ff40bfe [Ilya Ganelin] Updated tests to fix small bugs
19c31af [Ilya Ganelin] Added cleaner computation of time conversions in tests
6387772 [Ilya Ganelin] Updated suffix handling to handle overlap of units more gracefully
5193d5f [Ilya Ganelin] Resolved merge conflicts
76cfa27 [Ilya Ganelin] [SPARK-5931] Minor nit fixes'
bf779b0 [Ilya Ganelin] Special handling of overlapping usffixes for java
dd0a680 [Ilya Ganelin] Updated scala code to call into java
b2fc965 [Ilya Ganelin] replaced get or default since it's not present in this version of java
39164f9 [Ilya Ganelin] [SPARK-5931] Updated Java conversion to be similar to scala conversion. Updated conversions to clean up code a little using TimeUnit.convert. Added Unit tests
3b126e1 [Ilya Ganelin] Fixed conversion to US from seconds
1858197 [Ilya Ganelin] Fixed bug where all time was being converted to us instead of the appropriate units
bac9edf [Ilya Ganelin] More whitespace
8613631 [Ilya Ganelin] Whitespace
1c0c07c [Ilya Ganelin] Updated Java code to add day, minutes, and hours
647b5ac [Ilya Ganelin] Udpated time conversion to use map iterator instead of if fall through
70ac213 [Ilya Ganelin] Fixed remaining usages to be consistent. Updated Java-side time conversion
68f4e93 [Ilya Ganelin] Updated more files to clean up usage of default time strings
3a12dd8 [Ilya Ganelin] Updated host revceiver
5232a36 [Ilya Ganelin] [SPARK-5931] Changed default behavior of time string conversion.
499bdf0 [Ilya Ganelin] Merge branch 'SPARK-5931' of github.com:ilganeli/spark into SPARK-5931
9e2547c [Ilya Ganelin] Reverting doc changes
8f741e1 [Ilya Ganelin] Update JavaUtils.java
34f87c2 [Ilya Ganelin] Update Utils.scala
9a29d8d [Ilya Ganelin] Fixed misuse of time in streaming context test
42477aa [Ilya Ganelin] Updated configuration doc with note on specifying time properties
cde9bff [Ilya Ganelin] Updated spark.streaming.blockInterval
c6a0095 [Ilya Ganelin] Updated spark.core.connection.auth.wait.timeout
5181597 [Ilya Ganelin] Updated spark.dynamicAllocation.schedulerBacklogTimeout
2fcc91c [Ilya Ganelin] Updated spark.dynamicAllocation.executorIdleTimeout
6d1518e [Ilya Ganelin] Upated spark.speculation.interval
3f1cfc8 [Ilya Ganelin] Updated spark.scheduler.revive.interval
3352d34 [Ilya Ganelin] Updated spark.scheduler.maxRegisteredResourcesWaitingTime
272c215 [Ilya Ganelin] Updated spark.locality.wait
7320c87 [Ilya Ganelin] updated spark.akka.heartbeat.interval
064ebd6 [Ilya Ganelin] Updated usage of spark.cleaner.ttl
21ef3dd [Ilya Ganelin] updated spark.shuffle.sasl.timeout
c9f5cad [Ilya Ganelin] Updated spark.shuffle.io.retryWait
4933fda [Ilya Ganelin] Updated usage of spark.storage.blockManagerSlaveTimeout
7db6d2a [Ilya Ganelin] Updated usage of spark.akka.timeout
404f8c3 [Ilya Ganelin] Updated usage of spark.core.connection.ack.wait.timeout
59bf9e1 [Ilya Ganelin] [SPARK-5931] Updated Utils and JavaUtils classes to add helper methods to handle time strings. Updated time strings in a few places to properly parse time
