mikemccand commented on a change in pull request #610: LUCENE-8671: Load FST
off-heap if reader is not opened from an index writer
URL: https://github.com/apache/lucene-solr/pull/610#discussion_r266947394
##########
File path:
lucene/core/src/java/org/apache/lucene/codecs/blocktree/FieldReader.java
##########
@@ -91,7 +92,8 @@
clone.seek(indexStartFP);
// Initialize FST offheap if index is MMapDirectory and
// docCount != sumDocFreq implying field is not primary key
- if (clone instanceof ByteBufferIndexInput && this.docCount !=
this.sumDocFreq) {
+ isFSTOffHeap = clone instanceof ByteBufferIndexInput && ((this.docCount
!= this.sumDocFreq) || openedFromWriter == false);
Review comment:
Is the idea here that it's only `IndexWriter` that needs fast ID lookups
(since it uses this when deleting docs by ID field)? E.g. apps that open
`IndexReader` themselves (outside of `IndexWriter`) don't need fast ID lookups
by default?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
