Operator install error when pulling from registry.connect.redhat.com

[davivcgarcia]

Version

$ ./openshift-install version
./openshift-install v0.13.0

Platform (aws|libvirt|openstack):

AWS

What happened?

The cluster does not have proper authentication credentials (service account) to access Red Hat Connect Registry (partners). When trying to deploy the MondoDB Operator, from Operator Hub, but the install stays in InstallWaiting and then Error. Checking the operator pod, it is reporting ImagePullBackOff:

Failed to pull image "registry.connect.redhat.com/mongodb/enterprise-operator": rpc error: code = Unknown desc = unable to retrieve auth token: invalid username/password

What you expected to happen?

Operator being installed successfully.

How to reproduce it (as minimally and precisely as possible)?

1. On Operator Hub page, install MongoDB.
2. On openshift-operator project, check for the mongodb-enterprise-operator pod events.

Anything else we need to know?

The problem does not affect, for example, the Etcd or AMQ Streams operators because the container images are being pulled from different registries (quay.io and registry.access.redhat.com):

Successfully pulled image "quay.io/coreos/etcd-operator@sha256:c0301e4686c3ed4206e370b42de5a3bd2229b9fb4906cf85f3f30650424abec2"

Successfully pulled image "registry.access.redhat.com/amqstreams-1/amqstreams10-clusteroperator-openshift:1.0.0"

[e-minguez]

Try to regenerate your token from try.openshift.com. I did regenerate and now includes the registry.connect.redhat.com token as well.

[davivcgarcia]

Thanks @e-minguez, I downloaded again the pull-secret.txt from try.openshift.com and the file is way bigger than the previous one, including authentication credentials for multiple registries. Do you know if we have an operator to reapply the new content to the hosts without the need to redeploy the cluster?

[e-minguez]

Unfortunately I have no idea... but I think your best chance is to redeploy it.

[wking]

We pass the pull secret in here, so try updating the kube-system/coreos-pull-secret secret. I dunno if that's actively monitored though; you may need to cycle compute nodes to get them to notice (and update masters by hand, until we support cycling them too).

[abhinavdahiya]

We pass the pull secret in here, so try updating the kube-system/coreos-pull-secret secret. I dunno if that's actively monitored though; you may need to cycle compute nodes to get them to notice (and update masters by hand, until we support cycling them too).

It should, and if it doesn't try opening issue on https://github.com/openshift/machine-config-operator/issues/new

[abhinavdahiya]

/close

[openshift-ci-robot]

@abhinavdahiya: Closing this issue.

In response to this:

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.