rmuir commented on PR #11987: URL: https://github.com/apache/lucene/pull/11987#issuecomment-1332427293
Thanks, yeah my remaining concern is the non-optimized merge... especially for those that delete and update documents (as it prevents them from getting optimized merges). Alternative solution to this issue might be, instead of removing the reuse from Decompressor, to instead try removing the stored fields/term vectors CloseableThreadLocals from SegmentCoreReaders.... this is more difficult as we'd have to change APIs around IndexReader to no longer call `document(i)` but instead call e.g. `getFieldsReader()` and then pull documents from that. It might alleviate the pressure, while still allowing merge to reuse stuff efficiently and queries to reuse stuff efficiently when pulling their top N, but it would require bigger changes. cc @jpountz -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
