Fast path for reading single doc with ordinals #102902
Merged
Conversation
This file contains 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
the
Team:QL (Deprecated)
|
Hi @dnhatn, I've created a changelog YAML for you. |
|
Pinging @elastic/es-ql (Team:QL) |
|
Pinging @elastic/elasticsearch-esql (:Query Languages/ES|QL) |
dnhatn
added a commit
that referenced
this pull request
Dec 5, 2023
Requesting and returning memory from a CircuitBreaker can be costly due to the involvement of read/write on one or several atomic longs. To address this issue, the local breaker adopts a strategy of over-requesting memory, utilizing the reserved amount for subsequent memory requests without direct access to the actual breaker. Before passing a Block to another Driver, it is necessary to switch the owning block factory to its parent, which is associated with the global breaker. This is done to bypass the local breaker when releasing memory, as the releasing thread can be any thread, not necessarily the one executing the Driver. There are two specific operators that need to change the owning block factory: SinkOperator (superset of ExchangeSinkOperator), which is the last operator of a Driver, and AsyncOperator, which can be responded by any thread in response. The optimization reduces the latency of the enrich operation in the nyc_taxis benchmark from 100ms to 50ms. When combined with #102902, it further reduces the latency to below 40ms, better than the previous performance before the regression. Relates #102625
nik9000
approved these changes
Dec 6, 2023
|
Thank you, Nik! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
This optimization is added for enrich lookups, which are likely to match a single document. The change decreases the latency of the enrich operation in the nyc_taxis benchmark from 100ms to 70ms. When combined with #102901, it further reduces the latency to below 40ms, better than the previous performance before the regression.
Relates #102625