Fix assertion error on multiple node search reduce errors #131085
+31
−0
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
When receiving exceptions from more than one node on data node search reduction, SearchQueryThenFetchAsyncAction will raise more than one phase failure. This can lead to calling the listener in AbstractSearchAsyncAction more than once, which in turn in tests trips an assertion in ActionListener#assertFirstRun.
cbuescher
added
>test
elasticsearchmachine
added
the
Team:Search Foundations
|
Pinging @elastic/es-search-foundations (Team:Search Foundations) |
|
I ran into this issue while working on #129445 where another test in this suite fails for probably other reasons ("testSortMixedFieldTypes"). The test provokes IllegalArgumentExceptions when sorting on two indices with incompatible sort types. Without the change in I'm not sure preventing propagating any subsequent phase failure in AbstractSearchAsyncAction like suggested here is the right way to solve this, also I'm not really sure what big of an issue calling the ActionListener twice is in practice but the test assertion seems to be there for a reason. Also this might still be related to the weird errors we get in #129445 since that issue also seems to be related to search exception handling. |
benchaplin
reviewed
Jul 16, 2025
There was a problem hiding this comment.
I agree it seems wrong to raise multiple phase failures for multiple data node failures.
I'm hesitant about this change, is it possible we're relying on multiple runs of raisePhaseFailure / listener.onFailure(exception) here, one per node search exception? Would love to hear more of your thoughts on that.
| @@ -621,6 +622,10 @@ protected BytesReference buildSearchContextId(ShardSearchFailure[] failures) { | |||
| * @param cause the cause of the phase failure | |||
| */ | |||
| public void onPhaseFailure(String phase, String msg, Throwable cause) { | |||
| if (phaseFailureEncountered.compareAndSet(false, true) == false) { | |||
There was a problem hiding this comment.
Maybe this is what we want, but this change would be lighter weight if this check was moved to the caller in SearchQueryThenFetchAsyncAction.
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.
When receiving exceptions from more than one node on data node search reduction, SearchQueryThenFetchAsyncAction will raise more than one phase failure. This can lead to calling the listener in AbstractSearchAsyncAction more than once, which in turn in tests trips an assertion in ActionListener#assertFirstRun.