branch-4.0: [fix](cloud) Prevent tablet KV leak after partial recycle failure #63377

[wyxxxcat]

pick: #63377

[hello-stephen]

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?

[wyxxxcat]

run buildall

[hello-stephen]

Cloud UT Coverage Report

Increment line coverage 95.31% (61/64) 🎉

Increment coverage report
Complete coverage report

Category	Coverage
Function Coverage	78.56% (1854/2360)
Line Coverage	65.33% (33222/50851)
Region Coverage	65.72% (16398/24952)
Branch Coverage	56.35% (8753/15534)

[wyxxxcat]

/review

[github-actions]

Automated review complete for #65249 at head d3f6ba4.

Summary: requesting changes for one correctness issue in the partial tablet recycle failure path. The new cleanup logic is covered by a test-only executor mode, but the production cancellation path still returns before successful tablets have their KVs removed.

Critical checkpoints:

• Prompt/readiness: read the review prompt first. No required AGENTS.md files were listed, and no repo-local code-review SKILL.md was present after hidden-path search.
• Scope: reviewed all six changed files, surrounding recycler/executor control flow, current GitHub PR patch metadata, and existing review context. No existing inline comments were present to deduplicate against.
• Validation: static review plus git diff --check. I did not run cloud tests because this checkout has no existing build tree or prebuilt recycler_test binary.
• User focus: no additional user-provided review focus was supplied.

Subagent conclusions:

• optimizer-rewrite appended NO_CANDIDATES_INITIAL; no optimizer/rewrite or parallel join/aggregate issue was found.
• tests-session-config proposed TSC-001, which was independently verified and accepted as MAIN-001 for the inline comment below.
• Convergence round 1 ended with both live subagents replying NO_NEW_VALUABLE_FINDINGS for the same current ledger/comment set after MAIN-001 was accepted.

dismiss