YARN-11127. Potential deadlock in AsyncDispatcher caused by RMNodeImp… #4270
Conversation
This file contains hidden or 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 overall
This message was automatically generated. |
|
To make sure unittest pass, I submit f887c7a. But I think this unittest fail, was not introduced by this PR. I test the two unittest in our MAC on trunk branch, still fail. |
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
Many unit test failed, reason is 'Metrics source DelegationTokenSecretManagerMetrics already exists!', wait for #4289 to solve this problem. I solve this problem by below way on my own mac. |
|
🎊 +1 overall
This message was automatically generated. |
goiri
reviewed
Jun 4, 2022
...-client/src/test/java/org/apache/hadoop/yarn/client/api/impl/TestRMDeadLockTriggerByApp.java
Outdated
Show resolved
Hide resolved
| yarnCluster.start(); | ||
|
|
||
| // start rm client | ||
| yarnClient = YarnClient.createYarnClient(); |
There was a problem hiding this comment.
Sorry I don't know this code standards. Should I fix all member variable like pattern this.xxx?
...-client/src/test/java/org/apache/hadoop/yarn/client/api/impl/TestRMDeadLockTriggerByApp.java
Outdated
Show resolved
Hide resolved
| attemptId = appReport.getCurrentApplicationAttemptId(); | ||
| appAttempt = | ||
| yarnCluster.getResourceManager().getRMContext().getRMApps() | ||
| .get(attemptId.getApplicationId()).getCurrentAppAttempt(); |
...-client/src/test/java/org/apache/hadoop/yarn/client/api/impl/TestRMDeadLockTriggerByApp.java
Show resolved
Hide resolved
...-client/src/test/java/org/apache/hadoop/yarn/client/api/impl/TestRMDeadLockTriggerByApp.java
Outdated
Show resolved
Hide resolved
| // Avoid the EventHandlingThread struck forever | ||
| if (deadLock) { | ||
| LOG.info("Found dead lock, stop EventHandlingThread manually!"); | ||
| ((AsyncDispatcher) rm.getRMContext().getDispatcher()) |
There was a problem hiding this comment.
We should probably use fail() after this.
| if (deadLock) { | ||
| LOG.info("Found dead lock, stop EventHandlingThread manually!"); | ||
| ((AsyncDispatcher) rm.getRMContext().getDispatcher()) | ||
| .forceEventHandlingThreadStop(); |
There was a problem hiding this comment.
Should we do the forceEventHandlingThreadStop() in the tearDown?
| Assert.assertFalse("There is dead lock!", deadLock); | ||
|
|
||
| // join all thread | ||
| allocateThread.join(); |
There was a problem hiding this comment.
if the assert triggers, this won't join()
…l, SchedulerApplicationAttempt and RMAppImpl's lock contention.
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
|
@goiri Can you help me review this again? This patch have been applied on our cluster for long time. |
|
@goiri @slfan1989 The main reason for the deadlock described in this issue is that there are many types of locks in ResourceManager, and we do not have a clear rule to restrict the use of locks. I think this is an important topic. For me, I think we should have this rules: We can only try to acquire the lock in this order: Node --> App --> AppAttempt.But the lock cannot be acquired in the opposite direction. And in addition, there is no need to use write lock protect |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Description of PR
detail message see: https://issues.apache.org/jira/browse/YARN-11127
How was this patch tested?
uni-test in PR
For code changes: