[KYUUBI #5531][TEST] Fix flaky FlinkOperationOnYarnSuite by enlarging the max rows setting #5549
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
1. "$@" is a array, we want use string to compare. so update "$@" => "$*" 2. `tty` mean execute the command, we can use $(tty) replace it 3. param $# is a number, compare number should use -gt/-lt,not >/<
1. "$@" is a array, we want use string to compare. so update "$@" => "$*" 2. `tty` mean execute the command, we can use $(tty) replace it 3. param $# is a number, compare number should use -gt/-lt,not >/< 4. not sure the /bin/kyuubi line 63 'exit -1' need modify? so the directory bin only have a shellcheck note in /bin/kyuubi
# Conflicts: # bin/docker-image-tool.sh
Revert "fix_4186"
# Conflicts: # externals/kyuubi-spark-sql-engine/src/main/scala/org/apache/spark/kyuubi/SQLOperationListener.scala # externals/kyuubi-spark-sql-engine/src/main/scala/org/apache/spark/kyuubi/SparkConsoleProgressBar.scala
|
@davidyuan1223 thanks for your contribution, would you like to consider using https://dev.to/lydiahallie/cs-visualized-useful-git-commands-37p1 |
ok, wait the ci success i will try to reduce the commit |
link3280
approved these changes
Oct 27, 2023
There was a problem hiding this comment.
I think it's a reasonable change.
FWIW, The root cause is that we cannot easily determine if all tasks of a Flink job are running. The tests in the Flink project would invoke a rest client to check the task status, but I think it's overkill for us.
Let's see how's it going with the new max rows conf.
just like the comment #5531 (comment), actually, i also not sure the kyuubi's conf can solve the problem, maybe we also need to change the create sql about datagen's rows-per-second's value. |
|
could you change the PR title with more accurate words? "fix bug" is too generic which makes less sense for future explorers. |
davidyuan1223
changed the title
pan3793
changed the title
Codecov Report
@@ Coverage Diff @@
## master #5549 +/- ##
=============================================
+ Coverage 0.00% 61.40% +61.40%
- Complexity 0 23 +23
=============================================
Files 588 600 +12
Lines 33480 34316 +836
Branches 4405 4501 +96
=============================================
+ Hits 0 21072 +21072
+ Misses 33480 11109 -22371
- Partials 0 2135 +2135
... and 541 files with indirect coverage changes 📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
pan3793
approved these changes
Oct 27, 2023
ok, i will fix those questions in the future. |
|
Also, would you mind centralizing your analysis to the PR description? it will be included in the commit message, making it convenient for future explorers to understand the background of this change. |
of course, and i think we can add desc about kyuubi's flink max-rows conf. maybe there have someone makes the some problem like this bug, it can improve user to find the source problem quickly, what do you think? (maybe i need more time to manage the commit history, i have same problem in other pr, Since I'm used to idea's git tool, the time it takes me to deal with this problem may not be solved until tomorrow...) |
|
@davidyuan1223 thanks for your patience, the clean commit history is just something nice to have, not a big deal. Your analysis of the issue is the most valuable thing. Even though this PR only contains one line change, I consider escalating the task level from easy to medium (if no objection from others) according to your deep dive of the root cause. |
I am educated from your analysis left in #5531, I think it's good to inline those parts to the PR description |
I have updated the content of PR, what do you think? |
bowenliang123
approved these changes
Oct 27, 2023
yaooqinn
reviewed
Oct 27, 2023
...sql-engine/src/test/scala/org/apache/kyuubi/engine/flink/operation/FlinkOperationSuite.scala
Show resolved
Hide resolved
Thank you for your compliment. I am still a newbie in the field of big data. I have a lot to learn from kyuubi and you. |
pan3793
reviewed
Oct 29, 2023
...sql-engine/src/test/scala/org/apache/kyuubi/engine/flink/operation/FlinkOperationSuite.scala
Outdated
Show resolved
Hide resolved
yaooqinn
approved these changes
Oct 30, 2023
…uubi/engine/flink/operation/FlinkOperationSuite.scala Co-authored-by: Cheng Pan <[email protected]>
pan3793
pushed a commit
that referenced
this pull request
Oct 30, 2023
… the max rows setting ### _Why are the changes needed?_ #### 1. know about this pr When we execute flink(1.17+) test case, it may throw exception when the test case is `show/stop job`, the exception desc like this ``` - execute statement - show/stop jobs *** FAILED *** org.apache.kyuubi.jdbc.hive.KyuubiSQLException: Error operating ExecuteStatement: org.apache.flink.table.gateway.service.utils.SqlExecutionException: Could not stop job 4dece26857fab91d63fad1abd8c6bdd0 with savepoint for operation 9ed8247a-b7bd-4004-875b-61ba654ab3dd. at org.apache.flink.table.gateway.service.operation.OperationExecutor.lambda$callStopJobOperation$11(OperationExecutor.java:628) at org.apache.flink.table.gateway.service.operation.OperationExecutor.runClusterAction(OperationExecutor.java:716) at org.apache.flink.table.gateway.service.operation.OperationExecutor.callStopJobOperation(OperationExecutor.java:601) at org.apache.flink.table.gateway.service.operation.OperationExecutor.executeOperation(OperationExecutor.java:434) at org.apache.flink.table.gateway.service.operation.OperationExecutor.executeStatement(OperationExecutor.java:195) at org.apache.kyuubi.engine.flink.operation.ExecuteStatement.executeStatement(ExecuteStatement.scala:64) at org.apache.kyuubi.engine.flink.operation.ExecuteStatement.runInternal(ExecuteStatement.scala:56) at org.apache.kyuubi.operation.AbstractOperation.run(AbstractOperation.scala:171) at org.apache.kyuubi.session.AbstractSession.runOperation(AbstractSession.scala:101) at org.apache.kyuubi.session.AbstractSession.$anonfun$executeStatement$1(AbstractSession.scala:131) at org.apache.kyuubi.session.AbstractSession.withAcquireRelease(AbstractSession.scala:82) at org.apache.kyuubi.session.AbstractSession.executeStatement(AbstractSession.scala:128) at org.apache.kyuubi.service.AbstractBackendService.executeStatement(AbstractBackendService.scala:67) at org.apache.kyuubi.service.TFrontendService.ExecuteStatement(TFrontendService.scala:252) at org.apache.kyuubi.shade.org.apache.hive.service.rpc.thrift.TCLIService$Processor$ExecuteStatement.getResult(TCLIService.java:1557) at org.apache.kyuubi.shade.org.apache.hive.service.rpc.thrift.TCLIService$Processor$ExecuteStatement.getResult(TCLIService.java:1542) at org.apache.kyuubi.shade.org.apache.thrift.ProcessFunction.process(ProcessFunction.java:39) at org.apache.kyuubi.shade.org.apache.thrift.TBaseProcessor.process(TBaseProcessor.java:39) at org.apache.kyuubi.service.authentication.TSetIpAddressProcessor.process(TSetIpAddressProcessor.scala:36) at org.apache.kyuubi.shade.org.apache.thrift.server.TThreadPoolServer$WorkerProcess.run(TThreadPoolServer.java:286) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1[149](https://github.com/apache/kyuubi/actions/runs/6649714451/job/18068699087?pr=5501#step:8:150)) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:750) Caused by: java.util.concurrent.ExecutionException: java.util.concurrent.CompletionException: java.util.concurrent.CompletionException: org.apache.flink.runtime.checkpoint.CheckpointException: Checkpoint triggering task Source: tbl_a[6] -> Sink: tbl_b[7] (1/1) of job 4dece26857fab91d63fad1abd8c6bdd0 is not being executed at the moment. Aborting checkpoint. Failure reason: Not all required tasks are currently running. at java.util.concurrent.CompletableFuture.reportGet(CompletableFuture.java:357) at java.util.concurrent.CompletableFuture.get(CompletableFuture.java:1928) at org.apache.flink.table.gateway.service.operation.OperationExecutor.lambda$callStopJobOperation$11(OperationExecutor.java:617) ... 22 more Caused by: java.util.concurrent.CompletionException: java.util.concurrent.CompletionException: org.apache.flink.runtime.checkpoint.CheckpointException: Checkpoint triggering task Source: tbl_a[6] -> Sink: tbl_b[7] (1/1) of job 4dece26857fab91d63fad1abd8c6bdd0 is not being executed at the moment. Aborting checkpoint. Failure reason: Not all required tasks are currently running. at java.util.concurrent.CompletableFuture.encodeRelay(CompletableFuture.java:326) at java.util.concurrent.CompletableFuture.completeRelay(CompletableFuture.java:338) at java.util.concurrent.CompletableFuture.uniRelay(CompletableFuture.java:925) at java.util.concurrent.CompletableFuture$UniRelay.tryFire(CompletableFuture.java:913) at java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:488) at java.util.concurrent.CompletableFuture.completeExceptionally(CompletableFuture.java:1990) at org.apache.flink.runtime.rpc.akka.AkkaInvocationHandler.lambda$invokeRpc$1(AkkaInvocationHandler.java:260) at java.util.concurrent.CompletableFuture.uniWhenComplete(CompletableFuture.java:774) at java.util.concurrent.CompletableFuture$UniWhenComplete.tryFire(CompletableFuture.java:750) at java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:488) at java.util.concurrent.CompletableFuture.completeExceptionally(CompletableFuture.java:1990) at org.apache.flink.util.concurrent.FutureUtils.doForward(FutureUtils.java:1298) at org.apache.flink.runtime.concurrent.akka.ClassLoadingUtils.lambda$null$1(ClassLoadingUtils.java:93) at org.apache.flink.runtime.concurrent.akka.ClassLoadingUtils.runWithContextClassLoader(ClassLoadingUtils.java:68) at org.apache.flink.runtime.concurrent.akka.ClassLoadingUtils.lambda$guardCompletionWithContextClassLoader$2(ClassLoadingUtils.java:92) at java.util.concurrent.CompletableFuture.uniWhenComplete(CompletableFuture.java:774) at java.util.concurrent.CompletableFuture$UniWhenComplete.tryFire(CompletableFuture.java:750) at java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:488) at java.util.concurrent.CompletableFuture.completeExceptionally(CompletableFuture.java:1990) at org.apache.flink.runtime.concurrent.akka.AkkaFutureUtils$1.onComplete(AkkaFutureUtils.java:45) at akka.dispatch.OnComplete.internal(Future.scala:299) at akka.dispatch.OnComplete.internal(Future.scala:297) at akka.dispatch.japi$CallbackBridge.apply(Future.scala:224) at akka.dispatch.japi$CallbackBridge.apply(Future.scala:221) at org.apache.kyuubi.operation.AbstractOperation.run(AbstractOperation.scala:171) ... Cause: java.lang.RuntimeException: org.apache.flink.util.SerializedThrowable:java.util.concurrent.CompletionException: org.apache.flink.runtime.checkpoint.CheckpointException: Checkpoint triggering task Source: tbl_a[6] -> Sink: tbl_b[7] (1/1) of job 4dece26857fab91d63fad1abd8c6bdd0 is not being executed at the moment. Aborting checkpoint. Failure reason: Not all required tasks are currently running. at java.util.concurrent.CompletableFuture.encodeRelay(CompletableFuture.java:326) at java.util.concurrent.CompletableFuture.completeRelay(CompletableFuture.java:338) at java.util.concurrent.CompletableFuture.uniRelay(CompletableFuture.java:925) at java.util.concurrent.CompletableFuture$UniRelay.tryFire(CompletableFuture.java:913) at java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:488) at java.util.concurrent.CompletableFuture.completeExceptionally(CompletableFuture.java:1990) at org.apache.flink.runtime.rpc.akka.AkkaInvocationHandler.lambda$invokeRpc$1(AkkaInvocationHandler.java:260) at java.util.concurrent.CompletableFuture.uniWhenComplete(CompletableFuture.java:774) at java.util.concurrent.CompletableFuture$UniWhenComplete.tryFire(CompletableFuture.java:750) at java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:488) ... Cause: java.lang.RuntimeException: org.apache.flink.util.SerializedThrowable:org.apache.flink.runtime.checkpoint.CheckpointException: Checkpoint triggering task Source: tbl_a[6] -> Sink: tbl_b[7] (1/1) of job 4dece26857fab91d63fad1abd8c6bdd0 is not being executed at the moment. Aborting checkpoint. Failure reason: Not all required tasks are currently running. at org.apache.flink.runtime.checkpoint.DefaultCheckpointPlanCalculator.checkTasksStarted(DefaultCheckpointPlanCalculator.java:143) at org.apache.flink.runtime.checkpoint.DefaultCheckpointPlanCalculator.lambda$calculateCheckpointPlan$1(DefaultCheckpointPlanCalculator.java:105) at java.util.concurrent.CompletableFuture$AsyncSupply.run(CompletableFuture.java:[160](https://github.com/apache/kyuubi/actions/runs/6649714451/job/18068699087?pr=5501#step:8:161)4) at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.lambda$handleRunAsync$4(AkkaRpcActor.java:453) at org.apache.flink.runtime.concurrent.akka.ClassLoadingUtils.runWithContextClassLoader(ClassLoadingUtils.java:68) at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleRunAsync(AkkaRpcActor.java:453) at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleRpcMessage(AkkaRpcActor.java:218) at org.apache.flink.runtime.rpc.akka.FencedAkkaRpcActor.handleRpcMessage(FencedAkkaRpcActor.java:84) at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleMessage(AkkaRpcActor.java:[168](https://github.com/apache/kyuubi/actions/runs/6649714451/job/18068699087?pr=5501#step:8:169)) at akka.japi.pf.UnitCaseStatement.apply(CaseStatements.scala:24) ... ``` #### 2. what make the test case failed? If we want know the reason about the exception, we need to understand the process of flink executing stop job, the process line like this code space show(it's source is our bad test case, we can use this test case to solve similar problems) ``` 1. sql 1.1 create table tbl_a (a int) with ('connector' = 'datagen','rows-per-second'='10') 1.2 create table tbl_b (a int) with ('connector' = 'blackhole') 1.3 insert into tbl_b select * from tbl_a 2. start job: it will get 2 tasks abount source sink 3. show job: we can get job info 4. stop job(the main error): 4.1 stop job need checkpoint 4.2 start checkpoint, it need all task state is running 4.3 checkpoint can not get all task state is running, then throw the exception ``` Actually, in a normal process, it should not throw the exception, if this happens to your job, please check your kyuubi conf `kyuubi.session.engine.flink.max.rows`, it's default value is 1000000. But in the test case, we only the the conf's value is 10. It's the reason to make the error, this conf makes when we execute a stream query, it will cancel the when the limit is reached. Because flink's datagen is a streamconnector, so we can imagine, when we execute those sql, because our conf, it will make the sink task be canceled because the query reached 10. So when we execute stop job, flink checkpoint cannot get the tasks about this job is all in running state, then flink throw this exception. #### 3. how can we solve this problem? When your job makes the same exception, please make sure your kyuubi conf `kyuubi.session.engine.flink.max.rows`'s value can it meet your streaming query needs? Then changes the conf's value. close #5531 ### _How was this patch tested?_ - [ ] Add some test cases that check the changes thoroughly including negative and positive cases if possible - [ ] Add screenshots for manual tests if appropriate - [x] [Run test](https://kyuubi.readthedocs.io/en/master/contributing/code/testing.html#running-tests) locally before make a pull request ### _Was this patch authored or co-authored using generative AI tooling?_ No Closes #5549 from davidyuan1223/fix_flink_test_bug. Closes #5531 ce7fd79 [david yuan] Update externals/kyuubi-flink-sql-engine/src/test/scala/org/apache/kyuubi/engine/flink/operation/FlinkOperationSuite.scala dc3a4b9 [davidyuan] fix flink on yarn test bug 86a647a [davidyuan] fix flink on yarn test bug cbd4c0c [davidyuan] fix flink on yarn test bug 8b51840 [davidyuan] add common method to get session level config bcb0cf3 [davidyuan] Merge remote-tracking branch 'origin/master' 72e7aea [david yuan] Merge branch 'apache:master' into master 57ec746 [david yuan] Merge pull request #13 from davidyuan1223/fix 56b91a3 [yuanfuyuan] fix_4186 c8eb9a2 [david yuan] Merge branch 'apache:master' into master 2beccb6 [david yuan] Merge branch 'apache:master' into master 0925a4b [david yuan] Merge pull request #12 from davidyuan1223/revert-11-fix_4186 40e80d9 [david yuan] Revert "fix_4186" c83836b [david yuan] Merge pull request #11 from davidyuan1223/fix_4186 360d183 [david yuan] Merge branch 'master' into fix_4186 b616044 [yuanfuyuan] fix_4186 e244029 [david yuan] Merge branch 'apache:master' into master bfa6cbf [davidyuan1223] Merge branch 'apache:master' into master 16237c2 [davidyuan1223] Merge branch 'apache:master' into master c48ad38 [yuanfuyuan] remove the used blank lines 55a0a43 [xiaoyuandajian] Merge pull request #10 from xiaoyuandajian/fix-#4057 cb11935 [yuan] Merge remote-tracking branch 'origin/fix-#4057' into fix-#4057 86e4e1c [yuan] fix-#4057 info: modify the shellcheck errors file in ./bin 1. "$@" is a array, we want use string to compare. so update "$@" => "$*" 2. `tty` mean execute the command, we can use $(tty) replace it 3. param $# is a number, compare number should use -gt/-lt,not >/< 4. not sure the /bin/kyuubi line 63 'exit -1' need modify? so the directory bin only have a shellcheck note in /bin/kyuubi dd39efd [袁福元] fix-#4057 info: 1. "$@" is a array, we want use string to compare. so update "$@" => "$*" 2. `tty` mean execute the command, we can use $(tty) replace it 3. param $# is a number, compare number should use -gt/-lt,not >/< Lead-authored-by: davidyuan <[email protected]> Co-authored-by: david yuan <[email protected]> Co-authored-by: yuanfuyuan <[email protected]> Co-authored-by: yuan <[email protected]> Co-authored-by: davidyuan1223 <[email protected]> Co-authored-by: xiaoyuandajian <[email protected]> Co-authored-by: 袁福元 <[email protected]> Signed-off-by: Cheng Pan <[email protected]> (cherry picked from commit 26f614a) Signed-off-by: Cheng Pan <[email protected]>
|
Thanks, merged to master/1.8 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Why are the changes needed?
1. know about this pr
When we execute flink(1.17+) test case, it may throw exception when the test case is
show/stop job, the exception desc like this2. what make the test case failed?
If we want know the reason about the exception, we need to understand the process of flink executing stop job, the process line like this code space show(it's source is our bad test case, we can use this test case to solve similar problems)
Actually, in a normal process, it should not throw the exception, if this happens to your job, please check your kyuubi conf
kyuubi.session.engine.flink.max.rows, it's default value is 1000000. But in the test case, we only the the conf's value is 10.It's the reason to make the error, this conf makes when we execute a stream query, it will cancel the when the limit is reached. Because flink's datagen is a streamconnector, so we can imagine, when we execute those sql, because our conf, it will make the sink task be canceled because the query reached 10. So when we execute stop job, flink checkpoint cannot get the tasks about this job is all in running state, then flink throw this exception.
3. how can we solve this problem?
When your job makes the same exception, please make sure your kyuubi conf
kyuubi.session.engine.flink.max.rows's value can it meet your streaming query needs? Then changes the conf's value.close #5531
How was this patch tested?
Add some test cases that check the changes thoroughly including negative and positive cases if possible
Add screenshots for manual tests if appropriate
Run test locally before make a pull request
Was this patch authored or co-authored using generative AI tooling?
No