Skip to content

branch-4.1: [fix](workload group) Drain adaptive flush callbacks before pool teardown #68264 - #68326

Open
github-actions[bot] wants to merge 1 commit into
branch-4.1from
auto-pick-68264-branch-4.1
Open

github-actions[bot] wants to merge 1 commit into
branch-4.1from
auto-pick-68264-branch-4.1

Conversation

@github-actions

Copy link
Copy Markdown
Contributor

Cherry-picked from #68264

…down (#68264)

### What problem does this PR solve?

Adaptive flush callbacks can outlive the ThreadPool objects they borrow.
Workload group ID changes make teardown cancel a different registration
name, and timer cancellation can miss a rearmed timer or free its state
before an already-started callback acquires its mutex. These races can
crash the backend during workload group changes or shutdown.

- Save a stable, per-pool registration key and cancel it even when
adaptive adjustment has subsequently been disabled.
- Read the final timer ID under the callback mutex, then wait for brpc
to finish any running callback before releasing its state.
- Serialize registration, cancellation and stop; drain duplicate
registrations and reject registration after stop.
- Stop adaptive callbacks before backend pool teardown, and cancel
registrations on direct WG scheduler destruction.

### Release note

Fix backend crashes caused by adaptive flush callbacks accessing
released thread pools during workload group changes or backend shutdown.
@github-actions
github-actions Bot requested a review from yiguolei as a code owner September 21, 2026 08:29
@hello-stephen

Copy link
Copy Markdown
Contributor

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

  1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
  2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
  3. What features were added. Why was this function added?
  4. Which code was refactored and why was this part of the code refactored?
  5. Which functions were optimized and what is the difference before and after the optimization?

@hello-stephen

Copy link
Copy Markdown
Contributor

run buildall

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants