Fix IndexStatsIT#testFilterCacheStats #135390
Conversation
|
Pinging @elastic/es-distributed-indexing (Team:Distributed Indexing) |
tlrx
left a comment
There was a problem hiding this comment.
I'm not sure this is the right fix. Did you have a chance to run the reproducing test seed using the exact commit at the time of the failure? If so, I wonder if it is not "just" an issue with a random translog settings being set by the test framework?
Something like ESIntegTestCase#setRandomIndexTranslogSettings randomly sets an ASYNC durability...
| for (IndexService indexService : indexServices) { | ||
| for (IndexShard indexShard : indexService) { | ||
| // Wait for global checkpoint to stabilize first to avoid it changing while sync() is in progress. | ||
| assertBusy(() -> assertEquals(indexShard.getLocalCheckpoint(), indexShard.getLastKnownGlobalCheckpoint())); |
There was a problem hiding this comment.
I would avoid introducing more assertBusy if possible, it really slows down tests execution :(
|
I checked failing seeds and not all of them have that setting. Also ASYNC durability has a fairly high chance of being produced so we would see this failure way more often. |
See #124447.