Compute the doc range more efficiently when flushing doc block #14447
Conversation
github-project-automation
bot
moved this to Open
in OpenSearch Lucene & Core Performance Tracking
stefanvodita
left a comment
stefanvodita
left a comment
There was a problem hiding this comment.
This might still apply to Lucene103PostingsWriter. Should we rebase?
|
Thanks for updating the PR to include |
stefanvodita
left a comment
stefanvodita
left a comment
There was a problem hiding this comment.
Thank you for updating the 10.3 postings writer! I'm happy to merge that change, but I would rather not mix in the 10.1 change since that release has already gone out and we're unlikely to release a follow-up unless we find serious bugs.
Could you also add an entry under 10.3 in CHANGES.txt?
|
Sorry, I don't know if I was clear enough. I was asking if we can also revert the change to Thanks for adding the changelog entry! |
My fault, I forgot to revert. |
stefanvodita
left a comment
stefanvodita
left a comment
There was a problem hiding this comment.
Thank you @bugmakerrrrrr!
github-project-automation
bot
moved this from Open
to Merged
in OpenSearch Lucene & Core Performance Tracking
* A specialized Trie for Block Tree Index (#14333) * Fix OOM of TestTrie (#14488) * Compute the doc range more efficiently when flushing doc block (#14447) * Fix TestForTooMuchCloning (#14547) * Fix tests: too many calls to IndexInput.clone during merging (#14595) * Simplify rootCodeFp (#14596) --------- Co-authored-by: panguixin <[email protected]>
4 participants
Description
Today, we sum all the doc delta to compute the range when flushing doc block, which is equivalent to
lastDocID - level0LastDocID.