Clean up ShardFollowTasks for deleted indices #44702
Merged
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
tlrx
added
>bug
:Distributed Indexing/CCR
Collaborator
|
Pinging @elastic/es-distributed |
Member
Author
|
@elasticmachine run elasticsearch-ci/1 |
jasontedor
reviewed
Jul 23, 2019
Member
jasontedor
left a comment
jasontedor
left a comment
There was a problem hiding this comment.
Thanks for picking this up. I left a suggestion.
x-pack/plugin/ccr/src/main/java/org/elasticsearch/xpack/ccr/action/ShardFollowTaskCleaner.java
Show resolved
Hide resolved
Member
Author
|
Thanks @jasontedor, I applied your feedback. Can you please have another look? |
Member
|
Can you apply this to 6.8 as well? |
Member
Author
|
Thanks @jasontedor. I'll backport this change down the line to 6.8. |
tlrx
added a commit
to tlrx/elasticsearch
that referenced
this pull request
Jul 24, 2019
Deleting a follower index does not delete its ShardFollowTasks, potentially leaving many persistent tasks in the cluster that cannot be allocated on nodes and unnecessary fill the logs. This commit adds a cluster state listener (ShardFollowTaskCleaner) that completes (with a failure) any persistent task that refers to a non existent follower index. I think that this bug has been introduced by elastic#34404: before this change the task would have been completed as failed and removed from the cluster state.
tlrx
added a commit
to tlrx/elasticsearch
that referenced
this pull request
Jul 24, 2019
Deleting a follower index does not delete its ShardFollowTasks, potentially leaving many persistent tasks in the cluster that cannot be allocated on nodes and unnecessary fill the logs. This commit adds a cluster state listener (ShardFollowTaskCleaner) that completes (with a failure) any persistent task that refers to a non existent follower index. I think that this bug has been introduced by elastic#34404: before this change the task would have been completed as failed and removed from the cluster state.
This was referenced Jul 24, 2019
tlrx
added a commit
that referenced
this pull request
Jul 24, 2019
…ex (#44801) This commit unmutes and renames the test that failed on CI (#44796) after #44702 has been merged. This test assumes that follow stats still exist after a follower index has been deleted. The follow stats are based on persistent tasks, and since #44702 the persistent tasks of deleted following indices are now automatically cleaned up to avoid to bloat the cluster state. I don't think we should report any follow stats for deleted indices and I don't think that this test makes much sense now the tasks are cleaned up. This is why the test has been renamed. closes #44796
tlrx
added a commit
to tlrx/elasticsearch
that referenced
this pull request
Jul 24, 2019
…ex (elastic#44801) This commit unmutes and renames the test that failed on CI (elastic#44796) after elastic#44702 has been merged. This test assumes that follow stats still exist after a follower index has been deleted. The follow stats are based on persistent tasks, and since elastic#44702 the persistent tasks of deleted following indices are now automatically cleaned up to avoid to bloat the cluster state. I don't think we should report any follow stats for deleted indices and I don't think that this test makes much sense now the tasks are cleaned up. This is why the test has been renamed. closes elastic#44796
tlrx
added a commit
to tlrx/elasticsearch
that referenced
this pull request
Jul 24, 2019
…ex (elastic#44801) This commit unmutes and renames the test that failed on CI (elastic#44796) after elastic#44702 has been merged. This test assumes that follow stats still exist after a follower index has been deleted. The follow stats are based on persistent tasks, and since elastic#44702 the persistent tasks of deleted following indices are now automatically cleaned up to avoid to bloat the cluster state. I don't think we should report any follow stats for deleted indices and I don't think that this test makes much sense now the tasks are cleaned up. This is why the test has been renamed. closes elastic#44796
tlrx
added a commit
to tlrx/elasticsearch
that referenced
this pull request
Jul 24, 2019
…ex (elastic#44801) This commit unmutes and renames the test that failed on CI (elastic#44796) after elastic#44702 has been merged. This test assumes that follow stats still exist after a follower index has been deleted. The follow stats are based on persistent tasks, and since elastic#44702 the persistent tasks of deleted following indices are now automatically cleaned up to avoid to bloat the cluster state. I don't think we should report any follow stats for deleted indices and I don't think that this test makes much sense now the tasks are cleaned up. This is why the test has been renamed. closes elastic#44796
tlrx
added a commit
that referenced
this pull request
Jul 25, 2019
Deleting a follower index does not delete its ShardFollowTasks, potentially leaving many persistent tasks in the cluster that cannot be allocated on nodes and unnecessary fill the logs. This commit adds a cluster state listener (ShardFollowTaskCleaner) that completes (with a failure) any persistent task that refers to a non existent follower index. I think that this bug has been introduced by #34404: before this change the task would have been completed as failed and removed from the cluster state. Backport of #44702 and #44801 on 7.x
tlrx
added a commit
that referenced
this pull request
Jul 25, 2019
Deleting a follower index does not delete its ShardFollowTasks, potentially leaving many persistent tasks in the cluster that cannot be allocated on nodes and unnecessary fill the logs. This commit adds a cluster state listener (ShardFollowTaskCleaner) that completes (with a failure) any persistent task that refers to a non existent follower index. I think that this bug has been introduced by #34404: before this change the task would have been completed as failed and removed from the cluster state. Backport of #44702 and #44801 to 7.3
tlrx
added a commit
that referenced
this pull request
Jul 25, 2019
Deleting a follower index does not delete its ShardFollowTasks, potentially leaving many persistent tasks in the cluster that cannot be allocated on nodes and unnecessary fill the logs. This commit adds a cluster state listener (ShardFollowTaskCleaner) that completes (with a failure) any persistent task that refers to a non existent follower index. I think that this bug has been introduced by #34404: before this change the task would have been completed as failed and removed from the cluster state. Backport of #44702 and #44801 on 6.8
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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.
Deleting a follower index does not delete its
ShardFollowTasks, potentially leaving many persistent tasks in the cluster that cannot be allocated on nodes and unnecessary fill the logs. This pull request adds a cluster state listener (ShardFollowTaskCleaner) that completes (with a failure) any persistent task that refers to a non existent follower index.I think that this bug has been introduced by #34404: before this change the task would have been completed as failed and removed from the cluster state.