### What changes were proposed in this pull request?
This PR aims to prevent `HeartbeatReceiver` asks `Executor` to re-register blocker manager when the SparkContext is already stopped.
### Why are the changes needed?
Currently, `HeartbeatReceiver` blindly asks re-registration for the new heartbeat message.
However, when SparkContext is stopped, we don't need to re-register new block manager.
Re-registration causes unnecessary executors' logs and and a delay on job termination.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Pass the CIs with the newly added test case.
Closes#31373 from dongjoon-hyun/SPARK-34273.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR adds a new configuration, `spark.storage.blockManagerMasterDriverHeartbeatTimeoutMs`.
### Why are the changes needed?
Currently, it's a hard-coded `10 minutes`.
### Does this PR introduce _any_ user-facing change?
No. The default value is the same.
### How was this patch tested?
Pass the CIs.
Closes#31383 from dongjoon-hyun/SPARK-34278.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Use `count` to simplify `find + size(or length)` operation, it's semantically consistent, but looks simpler.
**Before**
```
seq.filter(p).size
```
**After**
```
seq.count(p)
```
### Why are the changes needed?
Code Simpilefications.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#31374 from LuciferYang/SPARK-34275.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Allow broadcast blocks to be put during decommissioning since migrations don't apply to them and they may be stored as part of job exec.
### Why are the changes needed?
Potential race condition.
### Does this PR introduce _any_ user-facing change?
Removal of race condition.
### How was this patch tested?
New unit test.
Closes#31298 from holdenk/SPARK-34193-torrentbroadcast-blockmanager-decommissioning-potential-race-condition.
Authored-by: Holden Karau <hkarau@apple.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Ensure that if a stage fails in the UI page, the corresponding error message can be displayed correctly.
### Why are the changes needed?
errormessage is not handled properly in JavaScript. If the 'at' is not exist, the error message on the page will be blank.
I made wochanges,
1. `msg.indexOf("at")` => `msg.indexOf("\n")`
As shows ablove, truncated at the 'at' position will result in a strange abstract of the error message. If there is a `\n` worit is more reasonable to truncate at the '\n' position.
2. If the `\n` does not exist check whether the msg is more than 100. If true, then truncate the display to avoid too long error message
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Manual test shows as belows, just a js change:
before modified:
after modified
Closes#31314 from akiyamaneko/error_message_display_empty.
Authored-by: neko <echohlne@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
1. Add back Maven enforcer for duplicate dependencies check
2. More strict check on Hadoop versions which support shaded client in `IsolatedClientLoader`. To do proper version check, this adds a util function `majorMinorPatchVersion` to extract major/minor/patch version from a string.
3. Cleanup unnecessary code
### Why are the changes needed?
The Maven enforcer was removed as part of #30556. This proposes to add it back.
Also, Hadoop shaded client doesn't work in certain cases (see [these comments](https://github.com/apache/spark/pull/30701#discussion_r558522227) for details). This strictly checks that the current Hadoop version (i.e., 3.2.2 at the moment) has good support of shaded client or otherwise fallback to old unshaded ones.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Existing tests.
Closes#31203 from sunchao/SPARK-33212-followup.
Lead-authored-by: Chao Sun <sunchao@apple.com>
Co-authored-by: Chao Sun <sunchao@apache.org>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Using a function like `.mkString` or `.getLines` directly on a `scala.io.Source` opened by `fromFile`, `fromURL`, `fromURI ` will leak the underlying file handle, this pr use the `Utils.tryWithResource` method wrap the `BufferedSource` to ensure these `BufferedSource` closed.
### Why are the changes needed?
Avoid file handle leak.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#31323 from LuciferYang/source-not-closed.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Redact event SparkListenerEnvironmentUpdate in log when its processing time exceeded logSlowEventThreshold
### Why are the changes needed?
Credentials could be exposed when its processing time exceeded logSlowEventThreshold
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Manually tested
Closes#31335 from warrenzhu25/34232.
Authored-by: Warren Zhu <warren.zhu25@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Compare the 3.1.1 API doc with the latest release version 3.0.1. Fix the following issues:
- Add missing `Since` annotation for new APIs
- Remove the leaking class/object in API doc
### Why are the changes needed?
Fix the issues in the Spark 3.1.1 release API docs.
### Does this PR introduce _any_ user-facing change?
Yes, API doc changes.
### How was this patch tested?
Manually test.
Closes#31271 from xuanyuanking/SPARK-34185.
Lead-authored-by: Yuanjian Li <yuanjian.li@databricks.com>
Co-authored-by: Hyukjin Kwon <gurwls223@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Update Avro dependency to version 1.10.1
### Why are the changes needed?
To catch up multiple improvements of Avro as well as fix security issues on transitive dependencies.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Since there were no API changes required we just run the tests
Closes#31232 from iemejia/SPARK-27733-avro-upgrade.
Authored-by: Ismaël Mejía <iemejia@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Add executor number check.
### Why are the changes needed?
The test `decommission workers ensure that shuffle output is regenerated even with shuffle service` assumes it has two executor and both of two tasks can execute concurrently.
The two tasks will execute serially if there only one executor. The result is test is unexpceted. E.g.
```
[info] 5 did not equal 4 Expected 4 tasks but got List(0:0:0:0-SUCCESS, 0:0:1:0-FAILED, 0:0:1:1-SUCCESS, 0:1:0:0-SUCCESS, 1:0:0:0-SUCCESS) (DecommissionWorkerSuite.scala:190)
```
The failed task due to the first task finished and decommission the worker.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Pass exists test.
Closes#31255 from ulysses-you/SPARK-34166.
Authored-by: ulysses-you <ulyssesyou18@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Check SPARK_TESTING as lazy val to avoid slow down when there are many environment variables
### Why are the changes needed?
If there are many environment variables, sys.env slows is very slow. As Utils.isTesting is called very often during Dataframe-Optimization, this can slow down evaluation very much.
An example for triggering the problem can be found in the bug ticket https://issues.apache.org/jira/browse/SPARK-34115
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
With the example provided in the ticket.
Closes#31244 from nob13/bug/34115.
Lead-authored-by: Norbert Schultz <norbert.schultz@reactivecore.de>
Co-authored-by: Norbert Schultz <noschultz@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
This PR aims to support fallback storage clean-up during stopping `SparkContext`.
### Why are the changes needed?
SPARK-33545 added `Support Fallback Storage during worker decommission` for the managed cloud-storages with TTL support. Usually, it's one day. This PR will add an additional clean-up feature during stopping `SparkContext` in order to save some money before TTL or the other HDFS-compatible storage which doesn't have TTL support.
### Does this PR introduce _any_ user-facing change?
Yes, but this is a new feature.
### How was this patch tested?
Pass the newly added UT.
Closes#31215 from dongjoon-hyun/SPARK-34142.
Authored-by: Dongjoon Hyun <dongjoon@apache.org>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Display history server entries using datatables instead of Mustache + Datatables which proved to be faster and non-blocking for the webpage while searching (using search bar in the page)
### Why are the changes needed?
Small changes in the attempts (entries) and removed part of HTML (Mustache template).
### Does this PR introduce _any_ user-facing change?
Not very sure, but it's not supposed to change the way the page looks rather it changes how entries are displayed.
### How was this patch tested?
Running test, since it's not adding new functionality.
Closes#31191 from mohan3d/feat/history-server-ui-optimization.
Lead-authored-by: mohan3d <mohan3d94@gmail.com>
Co-authored-by: Author: mohan3d <mohan3d94@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
This:
1. switches Spark to use shaded Hadoop clients, namely hadoop-client-api and hadoop-client-runtime, for Hadoop 3.x.
2. upgrade built-in version for Hadoop 3.x to Hadoop 3.2.2
Note that for Hadoop 2.7, we'll still use the same modules such as hadoop-client.
In order to still keep default Hadoop profile to be hadoop-3.2, this defines the following Maven properties:
```
hadoop-client-api.artifact
hadoop-client-runtime.artifact
hadoop-client-minicluster.artifact
```
which default to:
```
hadoop-client-api
hadoop-client-runtime
hadoop-client-minicluster
```
but all switch to `hadoop-client` when the Hadoop profile is hadoop-2.7. A side affect from this is we'll import the same dependency multiple times. For this I have to disable Maven enforcer `banDuplicatePomDependencyVersions`.
Besides above, there are the following changes:
- explicitly add a few dependencies which are imported via transitive dependencies from Hadoop jars, but are removed from the shaded client jars.
- removed the use of `ProxyUriUtils.getPath` from `ApplicationMaster` which is a server-side/private API.
- modified `IsolatedClientLoader` to exclude `hadoop-auth` jars when Hadoop version is 3.x. This change should only matter when we're not sharing Hadoop classes with Spark (which is _mostly_ used in tests).
### Why are the changes needed?
Hadoop 3.2.2 is released with new features and bug fixes, so it's good for the Spark community to adopt it. However, latest Hadoop versions starting from Hadoop 3.2.1 have upgraded to use Guava 27+. In order to resolve Guava conflicts, this takes the approach by switching to shaded client jars provided by Hadoop. This also has the benefits of avoid pulling other 3rd party dependencies from Hadoop side so as to avoid more potential future conflicts.
### Does this PR introduce _any_ user-facing change?
When people use Spark with `hadoop-provided` option, they should make sure class path contains `hadoop-client-api` and `hadoop-client-runtime` jars. In addition, they may need to make sure these jars appear before other Hadoop jars in the order. Otherwise, classes may be loaded from the other non-shaded Hadoop jars and cause potential conflicts.
### How was this patch tested?
Relying on existing tests.
Closes#30701 from sunchao/test-hadoop-3.2.2.
Authored-by: Chao Sun <sunchao@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Some local variables are declared as `var`, but they are never reassigned and should be declared as `val`, so this pr turn these from `var` to `val` except for `mockito` related cases.
### Why are the changes needed?
Use `val` instead of `var` when possible.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#31142 from LuciferYang/SPARK-33346.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
This pr use `exists` or `forall` to simplify `filter + emptiness check`, it's semantically consistent, but looks simpler. The rule as follow:
- `seq.filter(p).size == 0)` -> `!seq.exists(p)`
- `seq.filter(p).length > 0` -> `seq.exists(p)`
- `seq.filterNot(p).isEmpty` -> `seq.forall(p)`
- `seq.filterNot(p).nonEmpty` -> `!seq.forall(p)`
### Why are the changes needed?
Code Simpilefications.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#31184 from LuciferYang/SPARK-34118.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Fix a regression from https://github.com/apache/spark/pull/29959.
In Spark, the following file paths are considered as hidden paths and they are ignored on file reads:
1. starts with "_" and doesn't contain "="
2. starts with "."
However, after the refactoring PR https://github.com/apache/spark/pull/29959, the hidden paths are not filtered out on partition inference: https://github.com/apache/spark/pull/29959/files#r556432426
This PR is to fix the bug. To archive the goal, the method `InMemoryFileIndex.shouldFilterOut` is refactored as `HadoopFSUtils.shouldFilterOutPathName`
### Why are the changes needed?
Bugfix
### Does this PR introduce _any_ user-facing change?
Yes, it fixes a bug for reading file paths with partitions.
### How was this patch tested?
Unit test
Closes#31169 from gengliangwang/fileListingBug.
Authored-by: Gengliang Wang <gengliang.wang@databricks.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
There are some redundant collection conversion can be removed, for version compatibility, clean up these with Scala-2.13 profile.
### Why are the changes needed?
Remove redundant collection conversion
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
- Pass the Jenkins or GitHub Action
- Manual test `core`, `graphx`, `mllib`, `mllib-local`, `sql`, `yarn`,`kafka-0-10` in Scala 2.13 passed
Closes#31125 from LuciferYang/SPARK-34068.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
This pr use `exists` to simplify `find + emptiness check`, it's semantically consistent, but looks simpler.
**Before**
```
seq.find(p).isDefined
or
seq.find(p).isEmpty
```
**After**
```
seq.exists(p)
or
!seq.exists(p)
```
### Why are the changes needed?
Code Simpilefications.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#31130 from LuciferYang/SPARK-34070.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Add min threshold time speculation config
### Why are the changes needed?
When we turn on speculation with default configs we have the last 10% of the tasks subject to speculation. There are a lot of stages where the stage runs for few seconds to minutes. Also in general we don't want to speculate tasks that run within a minimum threshold. By setting a minimum threshold for speculation config gives us better control for speculative tasks
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Unit test
Closes#30710 from redsanket/SPARK-33741.
Lead-authored-by: schintap <schintap@verizonmedia.com>
Co-authored-by: Sanket Chintapalli <chintapalli.sanketreddy@gmail.com>
Signed-off-by: Thomas Graves <tgraves@apache.org>
### What changes were proposed in this pull request?
Increase test timeout.
### Why are the changes needed?
It's more reasonable to use 60s instead of 6s since many code place use the 60s in `TestUtils.waitUntilExecutorsUp`.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Pass exists test.
Closes#31166 from ulysses-you/SPARK-32850-FOLLOWUP.
Authored-by: ulysses-you <ulyssesyou18@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Add shouldInterruptTaskThread check when kill barrier task.
### Why are the changes needed?
We should interrupt task thread if user set local property `SPARK_JOB_INTERRUPT_ON_CANCEL` to true.
### Does this PR introduce _any_ user-facing change?
Yes, task will be interrupted if user set `SPARK_JOB_INTERRUPT_ON_CANCEL` to true.
### How was this patch tested?
Add test.
Closes#31127 from ulysses-you/SPARK-34069.
Authored-by: ulysses-you <ulyssesyou18@gmail.com>
Signed-off-by: Mridul Muralidharan <mridul<at>gmail.com>
### What changes were proposed in this pull request?
Missing POD detection is extended by timestamp (and time limit) based check to avoid wrongfully detection of missing POD detection.
The two new timestamps:
- `fullSnapshotTs` is introduced for the `ExecutorPodsSnapshot` which only updated by the pod polling snapshot source
- `registrationTs` is introduced for the `ExecutorData` and it is initialized at the executor registration at the scheduler backend
Moreover a new config `spark.kubernetes.executor.missingPodDetectDelta` is used to specify the accepted delta between the two.
### Why are the changes needed?
Watching a POD (`ExecutorPodsWatchSnapshotSource`) only inform about single POD changes. This could wrongfully lead to detecting of missing PODs (PODs known by scheduler backend but missing from POD snapshots) by the executor POD lifecycle manager.
A key indicator of this error is seeing this log message:
> "The executor with ID [some_id] was not found in the cluster but we didn't get a reason why. Marking the executor as failed. The executor may have been deleted but the driver missed the deletion event."
So one of the problem is running the missing POD detection check even when a single POD is changed without having a full consistent snapshot about all the PODs (see `ExecutorPodsPollingSnapshotSource`).
The other problem could be the race between the executor POD lifecycle manager and the scheduler backend: so even in case of a having a full snapshot the registration at the scheduler backend could precede the snapshot polling (and processing of those polled snapshots).
### Does this PR introduce _any_ user-facing change?
Yes. When the POD is missing then the reason message explaining the executor's exit is extended with both timestamps (the polling time and the executor registration time) and even the new config is mentioned.
### How was this patch tested?
The existing unit tests are extended.
Closes#30675 from attilapiros/SPARK-33711.
Authored-by: “attilapiros” <piros.attila.zsolt@gmail.com>
Signed-off-by: Holden Karau <hkarau@apple.com>
### What changes were proposed in this pull request?
For `AllJobsPage `class, `AllJobsPage` gets the schedulingMode of enumerated type by loading the `spark.scheduler.mode `configuration from Sparkconf, but an enumeration conversion error occurs when I set the value of this configuration to lowercase.
The reason for this problem is that the value of the `SchedulingMode `enumeration class is uppercase, which occurs when I configure `spark.scheduler.mode` to be lowercase.
I saw that the `#org.apache.spark.scheduler.TaskSchedulerImpl` class convert the `spark. scheduler.mode` value to uppercase, so I think it should be converted in `AllJobsPage `as well.
### Why are the changes needed?
An enumerated conversion error occurred with Spark when I set the value of this configuration to lowercase.
### How was this patch tested?
Existing tests.
Closes#31015 from yikf/master.
Authored-by: yikf <13468507104@163.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
According to https://github.com/apache/spark/pull/29966#discussion_r554514344
Use wrong name about suite file, this pr to fix this problem.
And change to use some fake ivy link for this test
### Why are the changes needed?
Follow file name rule
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
No
Closes#31118 from AngersZhuuuu/SPARK-33084-FOLLOW-UP.
Authored-by: angerszhu <angers.zhu@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
This PR takes over https://github.com/apache/spark/pull/28780.
1. Counted the spilled memory size when creating the `UnsafeExternalSorter` with the existing `InMemorySorter`
2. Accumulate the `totalSpillBytes` when merging two `UnsafeExternalSorter`
### Why are the changes needed?
As mentioned in https://github.com/apache/spark/pull/28780:
> It happends when hash aggregate downgrades to sort based aggregate.
`UnsafeExternalSorter.createWithExistingInMemorySorter` calls spill on an `InMemorySorter` immediately, but the memory pointed by `InMemorySorter` is acquired by outside `BytesToBytesMap`, instead the allocatedPages in `UnsafeExternalSorter`. So the memory spill bytes metric is always 0, but disk bytes spill metric is right.
Besides, this PR also fixes the `UnsafeExternalSorter.merge` by accumulating the `totalSpillBytes` of two sorters. Thus, we can report the correct spilled size in `HashAggregateExec.finishAggregate`.
Issues can be reproduced by the following step by checking the SQL metrics in UI:
```
bin/spark-shell --driver-memory 512m --executor-memory 512m --executor-cores 1 --conf "spark.default.parallelism=1"
scala> sql("select id, count(1) from range(10000000) group by id").write.csv("/tmp/result.json")
```
Before:
<img width="200" alt="WeChatfe5146180d91015e03b9a27852e9a443" src="https://user-images.githubusercontent.com/16397174/103625414-e6fc6280-4f75-11eb-8b93-c55095bdb5b8.png">
After:
<img width="200" alt="WeChat42ab0e73c5fbc3b14c12ab85d232071d" src="https://user-images.githubusercontent.com/16397174/103625420-e8c62600-4f75-11eb-8e1f-6f5e8ab561b9.png">
### Does this PR introduce _any_ user-facing change?
Yes, users can see the correct spill metrics after this PR.
### How was this patch tested?
Tested manually and added UTs.
Closes#31035 from Ngone51/SPARK-31952.
Lead-authored-by: yi.wu <yi.wu@databricks.com>
Co-authored-by: wangguangxin.cn <wangguangxin.cn@bytedance.com>
Signed-off-by: Wenchen Fan <wenchen@databricks.com>
### What changes were proposed in this pull request?
This PR is basically a followup of https://github.com/apache/spark/pull/14332.
Calling `map` alone might leave it not executed due to lazy evaluation, e.g.)
```
scala> val foo = Seq(1,2,3)
foo: Seq[Int] = List(1, 2, 3)
scala> foo.map(println)
1
2
3
res0: Seq[Unit] = List((), (), ())
scala> foo.view.map(println)
res1: scala.collection.SeqView[Unit,Seq[_]] = SeqViewM(...)
scala> foo.view.foreach(println)
1
2
3
```
We should better use `foreach` to make sure it's executed where the output is unused or `Unit`.
### Why are the changes needed?
To prevent the potential issues by not executing `map`.
### Does this PR introduce _any_ user-facing change?
No, the current codes look not causing any problem for now.
### How was this patch tested?
I found these item by running IntelliJ inspection, double checked one by one, and fixed them. These should be all instances across the codebase ideally.
Closes#31110 from HyukjinKwon/SPARK-34059.
Authored-by: HyukjinKwon <gurwls223@apache.org>
Signed-off-by: Liang-Chi Hsieh <viirya@gmail.com>
### What changes were proposed in this pull request?
This is the shuffle writer side change where executors can push data to remote shuffle services. This is needed for push-based shuffle - SPIP [SPARK-30602](https://issues.apache.org/jira/browse/SPARK-30602).
Summary of changes:
- This adds support for executors to push shuffle blocks after map tasks complete writing shuffle data.
- This also introduces a timeout specifically for creating connection to remote shuffle services.
### Why are the changes needed?
- These changes are needed for push-based shuffle. Refer to the SPIP in [SPARK-30602](https://issues.apache.org/jira/browse/SPARK-30602).
- The main reason to create a separate connection creation timeout is because the existing `connectionTimeoutMs` is overloaded and is used for connection creation timeouts as well as connection idle timeout. The connection creation timeout should be much lower than the idle timeouts. The default for `connectionTimeoutMs` is 120s. This is quite high for just establishing the connections. If a shuffle server node is bad then the connection creation will fail within few seconds. However, an overloaded shuffle server may take much longer to respond to a request and the channel can stay idle for a much longer time which is expected. Another reason is that with push-based shuffle, an executor may be fetching shuffle data and pushing shuffle data (next stage) simultaneously. Both these tasks will share the same connections with the shuffle service. If there is a bad shuffle server node and the connection creation timeout is very high then both these tasks end up waiting a long time time eventually impacting the performance.
### Does this PR introduce _any_ user-facing change?
Yes. This PR introduces client-side configs for push-based shuffle. If push-based shuffle is turned-off then the users will not see any change.
### How was this patch tested?
Added unit tests.
The reference PR with the consolidated changes covering the complete implementation is also provided in [SPARK-30602](https://issues.apache.org/jira/browse/SPARK-30602).
We have already verified the functionality and the improved performance as documented in the SPIP doc.
Lead-authored-by: Min Shen mshenlinkedin.com
Co-authored-by: Chandni Singh chsinghlinkedin.com
Co-authored-by: Ye Zhou yezhoulinkedin.com
Closes#30312 from otterc/SPARK-32917.
Lead-authored-by: Chandni Singh <singh.chandni@gmail.com>
Co-authored-by: Chandni Singh <chsingh@linkedin.com>
Co-authored-by: Min Shen <mshen@linked.in.com>
Co-authored-by: Ye Zhou <yezhou@linkedin.com>
Signed-off-by: Mridul Muralidharan <mridul<at>gmail.com>
### What changes were proposed in this pull request?
This PR makes `AppStatusListener` update the peak memory metrics for each Executor on task end like other peak memory metrics (e.g, stage, executors in a stage).
### Why are the changes needed?
When `AppStatusListener#onExecutorMetricsUpdate` is called, peak memory metrics for Executors, stages and executors in a stage are updated but currently, the metrics only for Executors are not updated on task end.
### Does this PR introduce _any_ user-facing change?
Yes. Executor peak memory metrics is updated more accurately.
### How was this patch tested?
After I run a job with `local-cluster[1,1,1024]` and visited `/api/v1/<appid>/executors`, I confirmed `peakExecutorMemory` metrics is shown for an Executor even though the life time of each job is very short .
I also modify the json files for `HistoryServerSuite`.
Closes#31029 from sarutak/update-executor-metrics-on-taskend.
Authored-by: Kousuke Saruta <sarutak@oss.nttdata.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Filter out the driver entity when updating the exclusion status of live executors(including the driver), so the UI won't be marked as excluded in the UI even if the node that hosts the driver has been marked as excluded.
### Why are the changes needed?
Before this change, if we run spark with the standalone mode and with spark.blacklist.enabled=true. The driver will be marked as excluded when the host that hosts that driver has been marked as excluded. While it's incorrect because the exclude list feature will exclude executors only and the driver is still active.
After the fix, the driver won't be marked as excluded.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Manual test. Reopen the UI and see the driver is no longer marked as excluded.
Closes#30954 from baohe-zhang/SPARK-33029.
Authored-by: Baohe Zhang <baohe.zhang@verizonmedia.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
From below log, Stage 600 could be removed from `stageAttemptToNumSpeculativeTasks` by `onStageCompleted()`, but the speculative task 306.1 in stage 600 threw `NoSuchElementException` when it entered into `onTaskEnd()`.
```
21/01/04 03:00:32,259 WARN [task-result-getter-2] scheduler.TaskSetManager:69 : Lost task 306.1 in stage 600.0 (TID 283610, hdc49-mcc10-01-0510-4108-039-tess0097.stratus.rno.ebay.com, executor 27): TaskKilled (another attempt succeeded)
21/01/04 03:00:32,259 INFO [task-result-getter-2] scheduler.TaskSetManager:57 : Task 306.1 in stage 600.0 (TID 283610) failed, but the task will not be re-executed (either because the task failed with a shuffle data fetch failure, so the
previous stage needs to be re-run, or because a different copy of the task has already succeeded).
21/01/04 03:00:32,259 INFO [task-result-getter-2] cluster.YarnClusterScheduler:57 : Removed TaskSet 600.0, whose tasks have all completed, from pool default
21/01/04 03:00:32,259 INFO [HiveServer2-Handler-Pool: Thread-5853] thriftserver.SparkExecuteStatementOperation:190 : Returning result set with 50 rows from offsets [5378600, 5378650) with 1fe245f8-a7f9-4ec0-bcb5-8cf324cbbb47
21/01/04 03:00:32,260 ERROR [spark-listener-group-executorManagement] scheduler.AsyncEventQueue:94 : Listener ExecutorAllocationListener threw an exception
java.util.NoSuchElementException: key not found: Stage 600 (Attempt 0)
at scala.collection.MapLike.default(MapLike.scala:235)
at scala.collection.MapLike.default$(MapLike.scala:234)
at scala.collection.AbstractMap.default(Map.scala:63)
at scala.collection.mutable.HashMap.apply(HashMap.scala:69)
at org.apache.spark.ExecutorAllocationManager$ExecutorAllocationListener.onTaskEnd(ExecutorAllocationManager.scala:621)
at org.apache.spark.scheduler.SparkListenerBus.doPostEvent(SparkListenerBus.scala:45)
at org.apache.spark.scheduler.SparkListenerBus.doPostEvent$(SparkListenerBus.scala:28)
at org.apache.spark.scheduler.AsyncEventQueue.doPostEvent(AsyncEventQueue.scala:38)
at org.apache.spark.scheduler.AsyncEventQueue.doPostEvent(AsyncEventQueue.scala:38)
at org.apache.spark.util.ListenerBus.postToAll(ListenerBus.scala:115)
at org.apache.spark.util.ListenerBus.postToAll$(ListenerBus.scala:99)
at org.apache.spark.scheduler.AsyncEventQueue.super$postToAll(AsyncEventQueue.scala:116)
at org.apache.spark.scheduler.AsyncEventQueue.$anonfun$dispatch$1(AsyncEventQueue.scala:116)
at scala.util.DynamicVariable.withValue(DynamicVariable.scala:62)
at org.apache.spark.scheduler.AsyncEventQueue.org$apache$spark$scheduler$AsyncEventQueue$$dispatch(AsyncEventQueue.scala:102)
at org.apache.spark.scheduler.AsyncEventQueue$$anon$2.$anonfun$run$1(AsyncEventQueue.scala:97)
at org.apache.spark.util.Utils$.tryOrStopSparkContext(Utils.scala:1320)
at org.apache.spark.scheduler.AsyncEventQueue$$anon$2.run(AsyncEventQueue.scala:97)
```
### Why are the changes needed?
To avoid throwing the java.util.NoSuchElementException
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
This is a protective patch and it's not easy to reproduce in UT due to the event order is not fixed in a async queue.
Closes#31025 from LantaoJin/SPARK-34000.
Authored-by: LantaoJin <jinlantao@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Fix un-correct doc of last change https://github.com/apache/spark/pull/30922#discussion_r551453193
### Why are the changes needed?
FIx doc
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Builds finished correctly.
Closes#31016 from AngersZhuuuu/SPARK-33908-FOLLOW-UP.
Authored-by: angerszhu <angers.zhu@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR upgrade Py4J from 0.10.9 to 0.10.9.1 that contains some bug fixes and improvements.
It contains one bug fix (4152353ac1).
### Why are the changes needed?
To leverage fixes from the upstream in Py4J.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Jenkins build and GitHub Actions will test it out.
Closes#31009 from HyukjinKwon/SPARK-33984.
Authored-by: HyukjinKwon <gurwls223@apache.org>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Check if the executorSummary.peakMemoryMetrics is defined before accessing it. Without checking, the UI has risked being stuck at the Executors page.
### Why are the changes needed?
App live UI may stuck at Executors page without this fix.
Steps to reproduce (with master branch):
In mac OS standalone mode, open a spark-shell
$SPARK_HOME/bin/spark-shell --master spark://localhost:7077
val x = sc.makeRDD(1 to 100000, 5)
x.count()
Then open the app UI in the browser, and click the Executors page, will get stuck at this page:
Also, the return JSON from API endpoint http://localhost:4040/api/v1/applications/app-20201224134418-0003/executors miss "peakMemoryMetrics" for executor objects. I attached the full json text in https://issues.apache.org/jira/browse/SPARK-33906.
I debugged it and observed that ExecutorMetricsPoller
.getExecutorUpdates returns an empty map, which causes peakExecutorMetrics to None in https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/status/LiveEntity.scala#L345. The possible reason for returning the empty map is that the stage completion time is shorter than the heartbeat interval, so the stage entry in stageTCMP has already been removed before the reportHeartbeat is called.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Manual test, rerun the steps of bug reproduce and see the bug is gone.
Closes#30920 from baohe-zhang/SPARK-33906.
Authored-by: Baohe Zhang <baohe.zhang@verizonmedia.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR proposed to make worker/executor decommission signal configurable.
* Added confs: `spark.worker.decommission.signal` / `spark.executor.decommission.signal`
* Rename `WorkerSigPWRReceived`/ `ExecutorSigPWRReceived` to `WorkerDecomSigReceived`/ `ExecutorDecomSigReceived`
### Why are the changes needed?
The current signal `PWR` can't work on macOS since it's not compliant with POSIX while macOS does. So the developers currently can't do end-to-end decommission test on their macOS environment.
Besides, the configuration becomes more flexible for users in case the default signal (`PWR`) gets conflicted with their own applications/environment.
### Does this PR introduce _any_ user-facing change?
No (it's a new API for 3.2)
### How was this patch tested?
Manually tested.
Closes#30968 from Ngone51/configurable-decom-signal.
Authored-by: yi.wu <yi.wu@databricks.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
There are only 3 compilation warnings related to `view bounds are deprecated` in Spark Code:
```
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/rdd/SequenceFileRDDFunctions.scala:35: view bounds are deprecated; use an implicit parameter instead.
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/rdd/SequenceFileRDDFunctions.scala:35: view bounds are deprecated; use an implicit parameter instead.
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/rdd/SequenceFileRDDFunctions.scala:55: view bounds are deprecated; use an implicit parameter instead.
```
This pr try to fix these compilation warnings.
### Why are the changes needed?
Fix compilation warnings about ` view bounds are deprecated`
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#30924 from LuciferYang/SPARK-33804.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Use the testing mode for the test to fix the flaky.
### Why are the changes needed?
The test is flaky:
```scala
[info] - SPARK-23365 Don't update target num executors when killing idle executors *** FAILED *** (126 milliseconds)
[info] 1 did not equal 2 (ExecutorAllocationManagerSuite.scala:1615)
[info] org.scalatest.exceptions.TestFailedException:
[info] at org.scalatest.Assertions.newAssertionFailedException(Assertions.scala:530)
[info] at org.scalatest.Assertions.newAssertionFailedException$(Assertions.scala:529)
[info] at org.scalatest.FunSuite.newAssertionFailedException(FunSuite.scala:1560)
[info] at org.scalatest.Assertions$AssertionsHelper.macroAssert(Assertions.scala:503)
[info] at org.apache.spark.ExecutorAllocationManagerSuite.$anonfun$new$84(ExecutorAllocationManagerSuite.scala:1617)
...
```
The root cause should be the same as https://github.com/apache/spark/pull/29773 since the test run under non-testing mode.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Manually checked. Flaky is gone by running the test hundreds of times after this fix.
Closes#30956 from Ngone51/fix-flaky-SPARK-23365.
Authored-by: yi.wu <yi.wu@databricks.com>
Signed-off-by: Wenchen Fan <wenchen@databricks.com>
### What changes were proposed in this pull request?
This is kind of a followup of https://github.com/apache/spark/pull/24033.
The first and last usage of that argument `SecurityManager` was removed in https://github.com/apache/spark/pull/24033.
After that, we don't need to pass `SecurityManager` anymore in `Utils.fetchFile` and related code paths.
This PR proposes to remove it out.
### Why are the changes needed?
For better readability of codes.
### Does this PR introduce _any_ user-facing change?
No, dev-only.
### How was this patch tested?
Manually complied. GitHub Actions and Jenkins build should test it out as well.
Closes#30945 from HyukjinKwon/SPARK-33925.
Authored-by: HyukjinKwon <gurwls223@apache.org>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR aims to fix offset bug and improve compression codec test coverage.
### Why are the changes needed?
When the user choose a non-default codec, it causes a failure.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Pass the extended test suite.
Closes#30934 from dongjoon-hyun/SPARK-33916.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Per discuss in https://github.com/apache/spark/pull/29966#discussion_r531917374
We'd better change `SparkSubmitUtils.resolveMavenCoordinates()` 's return value as `Seq[String]`
### Why are the changes needed?
refactor code
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Existed UT
Closes#30922 from AngersZhuuuu/SPARK-33908.
Authored-by: angerszhu <angers.zhu@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
There are total 15 compilation warnings about `Unicode escapes in triple quoted strings are deprecated` in Spark code now:
```
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2930: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2931: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2932: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2933: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2934: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2935: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2936: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/core/src/main/scala/org/apache/spark/util/Utils.scala:2937: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVExprUtils.scala:82: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/csv/CSVExprUtilsSuite.scala:32: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/csv/CSVExprUtilsSuite.scala:79: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/ParserUtilsSuite.scala:97: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/ParserUtilsSuite.scala:101: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/json/JsonParsingOptionsSuite.scala:76: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
[WARNING] /spark-source/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/json/JsonParsingOptionsSuite.scala:83: Unicode escapes in triple quoted strings are deprecated, use the literal character instead
```
This pr try to fix these warnnings.
### Why are the changes needed?
Cleanup compilation warnings about `Unicode escapes in triple quoted strings are deprecated`
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Pass the Jenkins or GitHub Action
Closes#30926 from LuciferYang/SPARK-33801.
Authored-by: yangjie01 <yangjie01@baidu.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>
### What changes were proposed in this pull request?
Support add jar with ivy path
### Why are the changes needed?
Since submit app can support ivy, add jar we can also support ivy now.
### Does this PR introduce _any_ user-facing change?
User can add jar with sql like
```
add jar ivy:://group:artifict:version?exclude=xxx,xxx&transitive=true
add jar ivy:://group:artifict:version?exclude=xxx,xxx&transitive=false
```
core api
```
sparkContext.addJar("ivy:://group:artifict:version?exclude=xxx,xxx&transitive=true")
sparkContext.addJar("ivy:://group:artifict:version?exclude=xxx,xxx&transitive=false")
```
#### Doc Update snapshot
### How was this patch tested?
Added UT
Closes#29966 from AngersZhuuuu/support-add-jar-ivy.
Lead-authored-by: angerszhu <angers.zhu@gmail.com>
Co-authored-by: AngersZhuuuu <angers.zhu@gmail.com>
Signed-off-by: Takeshi Yamamuro <yamamuro@apache.org>
### What changes were proposed in this pull request?
This PR aims to exclude fallback block manager from `executorList` function.
### Why are the changes needed?
When a fallback storage is used, the executors UI tab hangs because the executor list REST API result doesn't have `peakMemoryMetrics` of `ExecutorMetrics`. The root cause is that the block manager id used by fallback storage is included in the API result and it doesn't have `peakMemoryMetrics` because it's populated during HeartBeat reporting. We should hide it.
### Does this PR introduce _any_ user-facing change?
No. This is a bug fix on UI.
### How was this patch tested?
Manual. Run the following and visit Spark `executors` tab UI with browser.
```
bin/spark-shell -c spark.storage.decommission.fallbackStorage.path=file:///tmp/spark-storage/
```
Closes#30911 from dongjoon-hyun/SPARK-33893.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This is a retry of #30177.
This is not a complete fix, but it would take long time to complete (#30242).
As discussed offline, at least using `ContextAwareIterator` should be helpful enough for many cases.
As the Python evaluation consumes the parent iterator in a separate thread, it could consume more data from the parent even after the task ends and the parent is closed. Thus, we should use `ContextAwareIterator` to stop consuming after the task ends.
### Why are the changes needed?
Python/Pandas UDF right after off-heap vectorized reader could cause executor crash.
E.g.,:
```py
spark.range(0, 100000, 1, 1).write.parquet(path)
spark.conf.set("spark.sql.columnVector.offheap.enabled", True)
def f(x):
return 0
fUdf = udf(f, LongType())
spark.read.parquet(path).select(fUdf('id')).head()
```
This is because, the Python evaluation consumes the parent iterator in a separate thread and it consumes more data from the parent even after the task ends and the parent is closed. If an off-heap column vector exists in the parent iterator, it could cause segmentation fault which crashes the executor.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Added tests, and manually.
Closes#30899 from ueshin/issues/SPARK-33277/context_aware_iterator.
Authored-by: Takuya UESHIN <ueshin@databricks.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR aims to enable `spark.storage.replication.proactive` by default for Apache Spark 3.2.0.
### Why are the changes needed?
`spark.storage.replication.proactive` is added by SPARK-15355 at Apache Spark 2.2.0 and has been helpful when the block manager loss occurs frequently like K8s environment.
### Does this PR introduce _any_ user-facing change?
Yes, this will make the Spark jobs more robust.
### How was this patch tested?
Pass the existing UTs.
Closes#30876 from dongjoon-hyun/SPARK-33870.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR removed an unused variable `CompressionCodec.DEFAULT_COMPRESSION_CODEC`.
### Why are the changes needed?
Apache Spark 3.0.0 centralized this default value to `IO_COMPRESSION_CODEC.defaultValue` via [SPARK-26462](https://github.com/apache/spark/pull/23447).
We had better remove this variable to avoid any potential confusion in the future.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Pass the CI compilation.
Closes#30880 from dongjoon-hyun/minor.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR aims to test all compression codecs for encrypted spilling.
### Why are the changes needed?
To improve test coverage. Currently, only `CompressionCodec.DEFAULT_COMPRESSION_CODEC` is under testing.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Pass the CIs with the updated test cases.
Closes#30879 from dongjoon-hyun/SPARK-33873.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
