... just in case.
We already retry pushes of the image layers itself earlier in the pipeline. This last push should just be a tiny label push, but it can't hurt to retry it too.
Seems like you already have a solution for this, but I thought I'd offer a plug for a cool Python library that does something like this recently (and I packaged it for Fedora too!):
https://github.com/litl/backoff
LGTM, though it might be good to add a test to assert this.
The retry here is feature provided by Jenkins (https://jenkins.io/doc/pipeline/steps/workflow-basic-steps/#code-retry-code-retry-the-body-up-to-n-times).
retry
But the backoff library would be able to replace greenwave.utils.retry decorator (https://pagure.io/greenwave/c/251c008372a3a92f6b1f2a7ce164d064cf21a440).
backoff
greenwave.utils.retry
+1
Pull-Request has been merged by ralph
Thanks for this Ralph :-)
... just in case.
We already retry pushes of the image layers itself earlier in the pipeline.
This last push should just be a tiny label push, but it can't hurt to retry it
too.