jimczi commented on code in PR #14076:
URL: https://github.com/apache/lucene/pull/14076#discussion_r1888721064
##########
lucene/core/src/java/org/apache/lucene/codecs/lucene99/Lucene99FlatVectorsWriter.java:
##########
@@ -282,7 +285,7 @@ public CloseableRandomVectorScorerSupplier
mergeOneFieldToIndex(
// to perform random reads.
vectorDataInput =
segmentWriteState.directory.openInput(
- tempVectorData.getName(),
IOContext.DEFAULT.withReadAdvice(ReadAdvice.RANDOM));
+ tempVectorData.getName(),
IOContext.DEFAULT.withReadAdvice(readAdvice));
Review Comment:
It's used by any consumer of the vector scorer. This format is used
internally by HNSW and flat formats. See for instance Lucene99HnswVectorsFormat
where we set the read advice to random since HNSW is the main consumer. And
Lucene99ScalarQuantizedVectorsFormat where the read advice is set to sequential
since we only have the flat vectors available.
--
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]
