-
Notifications
You must be signed in to change notification settings - Fork 1.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Enhancement] Introduce pipeline-level multi-partitioned runtime filter(backport #33002) #31769
Open
satanson
wants to merge
1
commit into
StarRocks:branch-3.1
Choose a base branch
from
satanson:pipeline_level_rf
base: branch-3.1
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
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
satanson
force-pushed
the
pipeline_level_rf
branch
9 times, most recently
from
October 11, 2023 03:22
c7691ac
to
1b43bae
Compare
satanson
force-pushed
the
pipeline_level_rf
branch
from
October 17, 2023 18:07
1b43bae
to
54ebbc9
Compare
satanson
changed the title
pipeline-level rf
[Enhancement] Introduce pipeline-level multi-partitioned runtime filter
Oct 17, 2023
satanson
force-pushed
the
pipeline_level_rf
branch
3 times, most recently
from
October 18, 2023 00:39
1e723ce
to
4fa62f7
Compare
satanson
changed the title
[Enhancement] Introduce pipeline-level multi-partitioned runtime filter
[Enhancement] Introduce pipeline-level multi-partitioned runtime filter(backport #33002)
Oct 18, 2023
satanson
force-pushed
the
pipeline_level_rf
branch
2 times, most recently
from
October 23, 2023 02:59
e96360a
to
f218554
Compare
…er(backport StarRocks#33002) Signed-off-by: satanson <[email protected]>
satanson
force-pushed
the
pipeline_level_rf
branch
from
October 23, 2023 16:31
f218554
to
9e2619b
Compare
Kudos, SonarCloud Quality Gate passed! 0 Bugs 0.0% Coverage The version of Java (11.0.21) you have used to run this analysis is deprecated and we will stop accepting it soon. Please update to at least Java 17. |
[FE Incremental Coverage Report]❌ fail : 107 / 180 (59.44%) file detail
|
[BE Incremental Coverage Report]❌ fail : 12 / 113 (10.62%) file detail
|
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.
cherry-pick #33002
Pipeline-level runtime feature supports multi-partitioned pipeline-level rf, in the old-fashioned runtime filter construction, each driver in an instance just collects runtime filter raw materials and the driver who finishes HT constuction last constructs the singleton runtime filter finally, it is quite time-consuming when the right table of HashJoin is very large. In this feature, we eliminate this issue, each driver construct its own runtime filter, the last driver only merge the all of these runtime filters into a multi-partitioned runtime-filter with negligible cost.
This feature is controled by the session variable as follows:
We use a flatten array to hold multi-partitioned runtime filter for both global and instance rf, while runtime filter's layout is depends on distribution mode of HashJoin, pipeline_dop and local exchange interpolation, so its layout is not very clear, RuntimeFilterLayout infered from these information is introduced to describle the layout of runtime-filter.
RuntimeFilterLayout can describe both old-fashioned runtime filter and new-fashioned pipeline-level multi-partitioned runtime filter. its options as follows:
For old-fashioned runtime filter, we use options as following table shows
For the multi-partitioned pipeline-level rf, we use options as following table shows
What type of PR is this:
Does this PR entail a change in behavior?
If yes, please specify the type of change:
Checklist:
Bugfix cherry-pick branch check: