Skip to content

fix(smoke): report the failed range's recorded reason - #2488

Merged
Brad-Edwards merged 1 commit into
devfrom
fix/smoke-report-failure-reason
Oct 5, 2026
Merged

Brad-Edwards merged 1 commit into
devfrom
fix/smoke-report-failure-reason

Conversation

@Brad-Edwards

Copy link
Copy Markdown
Owner

Problem

When the post-deploy smoke's range fails, the smoke prints only range reached terminal status failed (request_id=...). The range's recorded failure reason lives only in the database, so diagnosing a failed smoke from CI needs in-cluster access (EKS API is private and pod logs are not shipped to CloudWatch).

Change

The smoke appends the range's recorded failure reason, bounded to 500 characters and sanitized, read through a new engine.services.get_range_failure_reason_by_request accessor (same shape as the existing by-request accessors).

Verification

tests/mission_control/test_run_post_deploy_smoke_command.py passes, including a new case that records a failure reason on a real range row and asserts it reaches the smoke's error.

When the post-deploy smoke's range failed, the smoke printed only
'range reached terminal status failed', and the range's recorded failure
reason lived only in the database, invisible from CI without cluster
access. The smoke now appends that reason, bounded and sanitized, read
through a new engine.services accessor (get_range_failure_reason_by_request).
@sonarqubecloud

sonarqubecloud Bot commented Oct 5, 2026

Copy link
Copy Markdown

@Brad-Edwards
Brad-Edwards merged commit 339a043 into dev Oct 5, 2026
80 checks passed
@Brad-Edwards
Brad-Edwards deleted the fix/smoke-report-failure-reason branch October 5, 2026 06:47
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.

1 participant