BryceKan3 commented on issue #15466:
URL: https://github.com/apache/lucene/issues/15466#issuecomment-3668484273
Hey this is an interesting edge case... The issue seems to be in the
createMultiSegmentIndex helper function of the test and not related to the code
itself. The test is creating 3 segments (or at least is supposed to...) _0.cfs
_1.cfs and _2.cfs. In this seed you sent its somehow forcing merges very
frequently (likely due to the low maxMBSortInHeap value and added bloom filter
postings files?) value which causes it to end up at one segment - _4.cfs where
when the test normally runs it doesn't do the merges.
```
{id=TestBloomFilteredLucenePostings(BloomFilteringPostingsFormat(PostingsFormat(name=Lucene104))),
content=FST50}, knn_vectors:{}, docValues:{}, maxPointsInLeafNode=1628,
maxMBSortInHeap=5.573857126302597,
sim=Asserting(RandomSimilarity(queryNorm=true): {content=IB LL-D2})
```
and then it doesn't actually hit the exception because there is only 1
segment so the reader gets opened without getting closed! The ideal fix here is
to ensure we are correctly creating multiple segments and preventing them from
getting merged which the test configuration you have seems to be causing.
<img width="2134" height="1024" alt="Image"
src="https://github.com/user-attachments/assets/7ac6e8b5-1c02-41fc-a012-f3c56838f668";
/>
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
