Use radix sort to speed up the sorting of deleted terms #12573
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gf2121
changed the title
jpountz
reviewed
Sep 21, 2023
Contributor
jpountz
left a comment
jpountz
left a comment
There was a problem hiding this comment.
Nice! I left some minor comments but it looks great in general.
lucene/core/src/java/org/apache/lucene/index/BufferedUpdates.java
Outdated
Show resolved
Hide resolved
lucene/core/src/java/org/apache/lucene/index/BufferedUpdates.java
Outdated
Show resolved
Hide resolved
jpountz
approved these changes
Sep 21, 2023
Contributor
jpountz
left a comment
jpountz
left a comment
There was a problem hiding this comment.
LGTM. I like that it also makes buffered deletes more memory-efficient as a side-effect.
Contributor
Author
|
Thanks @jpountz ! |
gf2121
added a commit
to gf2121/lucene
that referenced
this pull request
Sep 22, 2023
gf2121
added a commit
that referenced
this pull request
Sep 22, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Recently, we captured a flame graph in a scene with frequent updates, which showed that sorting deleted terms occupied a high CPU ratio.
In scenarios with many deleted terms, most terms could have the same field name. So a data structure like
Map<String, Map<BytesRef, Integer>>instead ofMap<Term, Integer>could be better here —— We can avoid field name compare and useMSBRadixSortto sort the bytes for each field.We can also take advantage of
BytesRefHashto implementMap<BytesRef, Integer>to get a more efficient memory layout, and there's already aMSBRadixSortimpl inBytesRefHashwe can reuse.Benchmark
We benchmarked the sort logic for 1,000,000 terms, showing 66% took decreasing:
An E2E benchmark that delete document with term 1,000,000 times showing took decreased 43% (with default iw config).