flake: TestRun_FirstPartDisarmsStartupTimeout

[flake-investigator]

CI Run: https://github.com/coder/coder/actions/runs/24164877546
Failed Job: test-go-pg-17 (job ID: 70524212886)
Commit: d9544603804c7cd4f6505a6b226c12e9623c7261 (author: Matt Vollmer)

Failure:

=== FAIL: coderd/x/chatd/chatloop TestRun_FirstPartDisarmsStartupTimeout (1.02s)
    chatloop_test.go:430:
        Error:      Not equal:
            expected: 1
            actual  : 2
    Test:       TestRun_FirstPartDisarmsStartupTimeout

Error analysis:

• The test expects only one attempt and no retry after the first stream part is received, but attempts incremented to 2 (startup timeout path retried).
• This suggests the startup timeout guard fired despite the first part being emitted, likely due to a timing race around disarming the startup timer vs the delayed delta parts.
• No data race warnings, panic traces, or OOM indicators found in the job logs.

Root cause classification: Flaky test (timing-sensitive startup timeout disarm race).

Assignment analysis:

• git log --oneline --follow coderd/x/chatd/chatloop/chatloop_test.go
• git show 70f031d793f0cde79f48acc2bd070a0abb7a1655 -- coderd/x/chatd/chatloop/chatloop_test.go
  • Commit 70f031d ("feat(coderd/chatd): structured chat error classification and retry hardening") added TestRun_FirstPartDisarmsStartupTimeout and the startup-timeout guard logic/tests.
• Assigning to @ethanndickson as the most recent non-trivial modifier of the failing test and startup-timeout behavior.

Related issues:

• No matching issues found after searching coder/internal for:
  • "TestRun_FirstPartDisarmsStartupTimeout"
  • "chatloop_test.go"
  • "startup timeout" + "chatloop"
  • "coderd/x/chatd"

Reproduction (likely flaky):

go test ./coderd/x/chatd/chatloop -run TestRun_FirstPartDisarmsStartupTimeout -count=1