-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
2.13.7
-
False
-
None
-
False
-
Not Started
-
Not Started
-
Not Started
-
Not Started
-
Not Started
-
Not Started
-
Workaround Exists
-
-
-
-
RHOAM Sprint 67
Expected behaviour
When following the documentation to perform the operator-based backup in a disconnected environment any images required should be able to be pulled from the mirror registry.
Current behaviour
When following the documentation the job fails due to this upstream image. It should be this image instead.
The backup job fails with:
Failed to pull image "quay.io/openshift/origin-cli:4.7": rpc error: code = DeadlineExceeded desc = pinging container registry quay.io: Get "https://quay.io/v2/": dial tcp [2600:1f18:483:cf01:27b5:90c9:eb73:3c17]:443: i/o timeout
UPDATE: The offending code is here and because the OC_CLI image is not set to the same as what midstream refers to then it falls back to this.
If we added the following function to the operator images file I believe that would be enough to fix it:
func OCCLIImageURL() string { return helper.GetEnvVar("RELATED_IMAGE_OC_CLI", component.OCCLIImageURL()) }
Because I can see in the deployment this image is being set under relatedImages already but due to the fact the code does not set the image to that env var then it can never use it.
relatedImages: <REDACTED ALL OTHER IMAGES> - image: >- registry.redhat.io/openshift4/ose-cli@sha256:1a8caae40be21f23ba3cb48b75809a73333907033a6041927aa568b5c9290f3c <- Set by midstream name: oc_cli