deardeng opened a new pull request, #60835:
URL: https://github.com/apache/doris/pull/60835

   Root cause:
   - In active-priority mode, performBalancing() returns early when active 
objects are still unbalanced. Under sustained hot-tablet pressure, this keeps 
happening and INACTIVE_ONLY never runs. Inactive partitions/tables are starved, 
so BE scale-out rebalancing becomes very slow.
   - runAfterCatalogReady() always called statRouteInfo() twice and rebuilt 
active tablet ids every round. These full-route/stat rebuilds allocate many 
temporary maps/sets, increasing young GC frequency.
   
   Fix:
   - Run the second statRouteInfo() only when smooth-upgrade migration actually 
moved tablets.
   - Add a starvation breaker: after N consecutive active-unbalanced rounds, 
force one INACTIVE_ONLY round.
   - Refresh active tablet snapshot by configurable interval (default 60s) 
instead of every round.
   
   Config:
   - cloud_active_tablet_ids_refresh_interval_second (default 60)
   - cloud_active_unbalanced_force_inactive_after_rounds (default 10)
   
   Test:
   - Add unit tests for forced inactive trigger, active snapshot refresh 
interval, and migrate no-op path.
   
   ### What problem does this PR solve?
   
   Issue Number: close #xxx
   
   Related PR: #xxx
   
   Problem Summary:
   
   ### Release note
   
   None
   
   ### Check List (For Author)
   
   - Test <!-- At least one of them must be included. -->
       - [ ] Regression test
       - [x] Unit Test
       - [ ] Manual test (add detailed scripts or steps below)
       - [ ] No need to test or manual test. Explain why:
           - [ ] This is a refactor/code format and no logic has been changed.
           - [ ] Previous test can cover this change.
           - [ ] No code files have been changed.
           - [ ] Other reason <!-- Add your reason?  -->
   
   - Behavior changed:
       - [x] No.
       - [ ] Yes. <!-- Explain the behavior change -->
   
   - Does this need documentation?
       - [x] No.
       - [ ] Yes. <!-- Add document PR link here. eg: 
https://github.com/apache/doris-website/pull/1214 -->
   
   ### Check List (For Reviewer who merge this PR)
   
   - [ ] Confirm the release note
   - [ ] Confirm test cases
   - [ ] Confirm document
   - [ ] Add branch pick label <!-- Add branch pick label that this PR should 
merge into -->
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to