[fix](fe) Clear warmup jobs on available backends

[bobhan1]

What problem does this PR solve?

Issue Number: None

Related PR: None

Problem Summary: CORE-6050 reports that canceling a cloud warm up job can fail to notify surviving BEs when client initialization fails on an unavailable BE. Keep the normal warmup RPC path fail-fast, but make CLEAR_JOB client initialization skip unavailable BEs and continue clearing reachable BEs.

Release note

None

Check List (For Author)

• Test: Unit Test
  • ./run-fe-ut.sh --run org.apache.doris.cloud.CloudWarmUpJobTest
• Behavior changed: Yes. Cloud warm up cancellation now best-effort clears reachable BEs even if another BE is unavailable.
• Does this need documentation: No

What problem does this PR solve?

Issue Number: close #xxx

Related PR: #xxx

Problem Summary:

Release note

None

Check List (For Author)

• Test

  • Regression test
  • 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

• Behavior changed:

  • No.
  • Yes.

• Does this need documentation?

  • No.
  • Yes.

Check List (For Reviewer who merge this PR)

• Confirm the release note
• Confirm test cases
• Confirm document
• Add branch pick label

[bobhan1]

run buildall

[bobhan1]

run nonConcurrent

[gavinchou]

/review

[github-actions]

Review result: no blocking issues found.

Critical checkpoint conclusions:

• Goal and proof: The PR keeps normal warm-up RPC client initialization fail-fast while making CLEAR_JOB best-effort across reachable BEs. The implementation matches that goal, and the added unit tests cover both the fail-fast SET_JOB path and the skip-unavailable CLEAR_JOB path.
• Scope and clarity: The change is small and localized to CloudWarmUpJob client initialization plus targeted unit tests.
• Concurrency and locking: cancel() and run() remain synchronized; the change does not add new shared state or new lock ordering. Existing RPCs still happen under the same synchronized job-level flow as before.
• Lifecycle: Borrowed clients are returned on successful paths and invalidated/removed on failed clear RPC paths; partially initialized normal warm-up clients are released on failure.
• Configuration/compatibility: No new configuration, storage format, thrift protocol, or FE-BE compatibility change.
• Parallel paths: The PR intentionally preserves fail-fast behavior for SET_JOB/normal warm-up and changes only CLEAR_JOB cleanup behavior.
• Error handling: Initialization failures during clear are logged per BE and do not block clearing other BEs; normal init still propagates failure.
• Tests: Added FE unit coverage is relevant. I attempted ./run-fe-ut.sh --run org.apache.doris.cloud.CloudWarmUpJobTest, but this runner could not generate sources because thirdparty/installed/bin/protoc was missing.
• Observability: New warning log includes BE id/address and job id for skipped clear initialization failures.
• Transaction/persistence/data correctness: No transaction, persistence, version visibility, or data-write path changes.
• Performance: No new hot-path loops or avoidable expensive work beyond existing per-BE client initialization/RPC behavior.

User focus: No additional user-provided review focus was present.

[github-actions]

PR approved by at least one committer and no changes requested.

[github-actions]

PR approved by anyone and no changes requested.