Since OCP 4.11, we have observed deployment timeout when installing OCP through a proxy.

In the ShiftStack QE environment, this issue is even more reoccurring with OCP 4.12 version and reproducing in ~90% of the OCP 4.12 proxy installations.

This issue was reported at - https://bugzilla.redhat.com/show_bug.cgi?id=2074497.

We suspect this issue is more reoccurring with newer OCP versions, as the container images pull rate grows along the versions.

The workaround for this issue is using the openshift-install flag to wait until the cluster is ready: [0]

openshift-install wait-for install-complete --log-level debug 

The same workaround is suggested for use in bare-metal workers installation. [1]

Can we document this issue and the workaround in OCP proxy installations on OpenStack? The Bug 2074497 has been open for a while.

Thanks!

[0] https://bugzilla.redhat.com/show_bug.cgi?id=2074497#c9

[1] https://github.com/openshift/installer/blob/master/docs/user/openstack/deploy_baremetal_workers.md#considerations-when-deploying-bare-metal-workers