Push to IBM cloud registry failing intermittently, retry if available can solve the issue.

[rashmispai]

Running kaniko build in IBM Cloud and we have builds happening for multiple components in parallel. Intermittently one of the components fail to push the container registry, this is causing us to do the complete pipeline again.

Whenever it fails, log is as below. A retry should obviously solve our problem but I see retry option is not yet available in Kaniko, any suggestion or help on solving this issue is much appreciated.

INFO[0084] Taking snapshot of files...
2020/05/20 10:47:51 existing blob: sha256:cca090ad6fefa18963f62e9cfb2c828e7486606e971dad7f8677040b39b1812d
error pushing image: failed to push to destination us.icr.io/test-pr/xyz01/allinoneappserver:extensions-20200519-1608: Head https://us.icr.io/v2/test-pr/xyz01/allinoneappserver/blobs/sha256:cb9162bf2ed92694f8f5b8fe9d55809651fe05ef55ae3bd231bde52b78b21fef: dial tcp: lookup us.icr.io on 172.21.0.10:53: no such host

Expected behavior

Retry to see if the image can be pushed.

Thanks for this great tool and expecting a solution and feedback to the issue.