Gracefully handle API call failure in cluster-events #1285
Conversation
There was a problem hiding this comment.
Would it be better from a separation-of-concerns perspective to simply allow this module to have the same EMPTY status message as any other module but indicate that there were API errors in the status of the upstream module? if we are already doing that then it should be adequate unless we want to consider preventing downstream modules from proceeding in that scenario . . . What's our established pattern in that respect?
An alternative to consider would be to automatically check for relevant records in ${erroredBronzeEventsTarget.tableFullName} at runtime rather than expecting the user to notice this advice in the pipReport.
There was a problem hiding this comment.
Please see this comment on the PR.
|
Gracefully handle API call failure in cluster-events
Solve #1284