Skip to content

Isolate query failures in publish pipeline (#45)#103

Open
BWilliams1004 wants to merge 1 commit into
mainfrom
bwilliams/45-query-failure-isolation
Open

Isolate query failures in publish pipeline (#45)#103
BWilliams1004 wants to merge 1 commit into
mainfrom
bwilliams/45-query-failure-isolation

Conversation

@BWilliams1004

Copy link
Copy Markdown

Closes #45. Catches exceptions in execute_and_upload_pg so a single failing query gets logged via Prefect's run logger instead of crashing the whole pipeline run. Since the page-building step already discovers files via an S3 glob rather than by inspecting task results, a failed query is simply skipped rather than blocking the archive/docs build for everything else.

The parquet file uploads fine, but the CSV write throws. You would end up with a partially published file even though the task reports a total failure.

Note on testing: Verified locally that this catches and logs exceptions correctly, but couldn’t run a full end-to-end test since I don’t yet have a working Prefect API connection (separate issue, following up on that). Would appreciate a check from someone with that access.

@BWilliams1004 BWilliams1004 requested a review from gankoji July 13, 2026 18:09
@BWilliams1004 BWilliams1004 self-assigned this Jul 13, 2026
@BWilliams1004 BWilliams1004 added the bug Something isn't working label Jul 13, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Make it so if a Query Fails, the rest will run.

1 participant