Block encode doc ids #134170
Merged
Block encode doc ids #134170
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
elasticsearchmachine
added
v9.2.0
needs:triage
benwtrent
added
>non-issue
:Search Relevance/Vectors
elasticsearchmachine
added
the
Team:Search Relevance
Collaborator
|
Pinging @elastic/es-search-relevance (Team:Search Relevance) |
tteofili
approved these changes
Sep 5, 2025
Contributor
tteofili
left a comment
tteofili
left a comment
There was a problem hiding this comment.
this seems reasonable, especially when running in memory constrained setups.
I would eventually change the bulk size in a different PR, to eventually isolate performance impacts.
iverase
approved these changes
Sep 5, 2025
Contributor
iverase
left a comment
iverase
left a comment
There was a problem hiding this comment.
LGTM
The reason for using 16 is that we use to filter soar assignments eagerly so it made sense to make the bulk size small.
benwtrent
added
the
auto-merge-without-approval
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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Outdated suggestions cannot be applied.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.
To protect against accidentally very large clusters, we shouldn't eagerly load ALL docIds for a postings list at time. This block encodes them into
BLOCK_SIZEchunks (16).Running benchmarks I found no measurable impact on index time. There MAY be a small increase at query time? The numbers aren't clear. If there is an impact, it seems very small.
As an aside, I am wondering if our block size is too conservative and if we should up it to 32 for vectors and docs as our typical running hardware is avx256 and 512...