Skip to content
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

*: adaptively control analyze distsql concurrency #53261

Conversation

hawkingrei
Copy link
Member

@hawkingrei hawkingrei commented May 14, 2024

What problem does this PR solve?

Issue Number: close #53262

Problem Summary:

What changed and how does it work?

When tidb cluster is larger and larger, the concurrency of the analyze scan should be larger to have a better performance. so we should make analyze_distsql_scan_concurrency adaptive.

BTW, Even a fourfold increase in concurrency seems conservative; for some customers, the actual concurrency is a thousand times the default value.

Check List

Tests

  • Unit test
  • Integration test
  • Manual test (add detailed scripts or steps below)
  • No need to test
    • I checked and no code files have been changed.

Side effects

  • Performance regression: Consumes more CPU
  • Performance regression: Consumes more Memory
  • Breaking backward compatibility

Documentation

  • Affects user behaviors
  • Contains syntax changes
  • Contains variable changes
  • Contains experimental features
  • Changes MySQL compatibility

Release note

Please refer to Release Notes Language Style Guide to write a quality release note.

None

…b node

Signed-off-by: Weizhen Wang <wangweizhen@pingcap.com>
Copy link

codecov bot commented May 14, 2024

Codecov Report

Attention: Patch coverage is 20.51282% with 31 lines in your changes are missing coverage. Please review.

Project coverage is 74.0185%. Comparing base (041f700) to head (d1fca57).
Report is 57 commits behind head on master.

Additional details and impacted files
@@               Coverage Diff                @@
##             master     #53261        +/-   ##
================================================
- Coverage   74.3906%   74.0185%   -0.3722%     
================================================
  Files          1493       1524        +31     
  Lines        356315     444505     +88190     
================================================
+ Hits         265065     329016     +63951     
- Misses        71913      95362     +23449     
- Partials      19337      20127       +790     
Flag Coverage Δ
integration 49.3124% <20.5128%> (?)
unit 71.5069% <20.5128%> (-1.6444%) ⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components Coverage Δ
dumpling 53.9957% <ø> (-2.3014%) ⬇️
parser ∅ <ø> (∅)
br 42.1070% <ø> (-0.8639%) ⬇️

Comment on lines 2752 to 2754
if concurrency <= 0 {
concurrency = adaptiveAnlayzeDistSQLConcurrency(context.Background(), b.ctx)
}
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could we merge this logic with AnalyzeDistSQLScanConcurrency or create a new function, since they are always called together?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done, it has been merged.

)

func adaptiveAnlayzeDistSQLConcurrency(ctx context.Context, sctx sessionctx.Context) int {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How often is this function called? Could it be invoked so frequently in some cases that it overwhelms the PD API server?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Every time, analyze jobs need to run.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How often is this function called? Could it be invoked so frequently in some cases that it overwhelms the PD API server?

Auto analyze. But I think it's QPS <= 10 qps

Signed-off-by: Weizhen Wang <wangweizhen@pingcap.com>
Copy link

ti-chi-bot bot commented May 21, 2024

[LGTM Timeline notifier]

Timeline:

  • 2024-05-17 11:10:51.218598398 +0000 UTC m=+1824404.975733971: ☑️ agreed by winoros.
  • 2024-05-21 05:23:17.694656768 +0000 UTC m=+2149151.451792338: ☑️ agreed by JmPotato.

@easonn7
Copy link

easonn7 commented May 21, 2024

/approve

@ti-chi-bot ti-chi-bot bot added the approved label May 21, 2024
@hawkingrei
Copy link
Member Author

/hold

@hi-rustin PTAL

@ti-chi-bot ti-chi-bot bot added the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label May 21, 2024
@hawkingrei
Copy link
Member Author

/retest

Copy link

ti-chi-bot bot commented May 21, 2024

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: easonn7, hi-rustin, JmPotato, winoros

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

@hawkingrei
Copy link
Member Author

/unhold

@ti-chi-bot ti-chi-bot bot removed the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label May 21, 2024
@hawkingrei
Copy link
Member Author

/retest

1 similar comment
@hawkingrei
Copy link
Member Author

/retest

@ti-chi-bot ti-chi-bot bot merged commit def7c23 into pingcap:master May 21, 2024
23 checks passed
RidRisR pushed a commit to RidRisR/tidb that referenced this pull request May 23, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
approved lgtm release-note-none size/M Denotes a PR that changes 30-99 lines, ignoring generated files.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

support adaptive analyze distsql scan concurreny
5 participants