Skip to content
This repository was archived by the owner on Aug 31, 2022. It is now read-only.

Reload BigQuery job after it is done in order to correctly capture errors #211

Merged
merged 1 commit into from
Dec 4, 2019
Merged

Reload BigQuery job after it is done in order to correctly capture errors #211

merged 1 commit into from
Dec 4, 2019

Conversation

jswarburton
Copy link
Contributor

In the case where a BigQuery job fails immediately (e.g. bucket does not exist in GCS) the job is never reloaded, and errors are lost.
In this case the job is treated as DONE, whilst the returned JobInfo may still report that the job is RUNNING.
By reloading the job after it is DONE, errors are captured and the true state of the job is returned from the library.

Hey, I just made a Pull Request!

Description

Motivation and Context

Have you tested this? If so, how?

Checklist for PR author(s)

  • Changes are covered by unit test
  • All tests pass
  • Code coverage check passes
  • Error handling is tested
  • Errors are handled at the appropriate layer
  • Errors that cannot be handled where they occur are propagated
  • Relevant documentation updated
  • This PR has NO breaking change to public API
  • This PR has breaking change to public API and it is documented

Checklist for PR reviewer(s)

  • This PR has been incorporated in release note for the coming version
  • Risky changes introduced by this PR have been all considered

…rors.

In the case where a BigQuery job fails immediately (e.g. bucket does not exist in GCS) the job is never reloaded, and errors are lost.
In this case the job is treated as DONE, whilst the returned JobInfo may still report that the job is RUNNING.
By reloading the job after it is DONE, errors are captured and the true state of the job is returned from the library.
@narape narape merged commit 98d3be4 into spotify:master Dec 4, 2019
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants