Let's say something like this is being returned when attempted to deploy an app to OpenShift, or when attempting to restart a deployment (scale up), or when restarting the OpenShift nodes.

Failed to pull image "registry.example.com/app001:latest": 
Error reading manifest latest in registry.example.com/app001
manifest unknown

If the pod exists, the oc describe pod command can be used to display the pod events. In this example, the events for pod001 will be displayed.

oc describe pod/pod001

Something like this may be returned.

Events:
  Type    Reason     Age   From                 Message
  ----    ------     ----  ----                 -------
  Normal  Pulling    3m    kubelet, os.example.com  Pulling image "repo.example.com/foo/bar:latest"
  Normal  Pulled     3m    kubelet, os.example.com  Failed to pull image "repo.example.com/foo/bar:latest"

Or something like this.

Events:
  Type     Reason   Age                      From     Message
  ----     ------   ----                     ----     -------
  Normal   Pulling  66m (x73 over 6h56m)     kubelet  Pulling image "registry.example.com/foo/bar:latest"
  Normal   BackOff  12m (x1767 over 6h55m)   kubelet  Back-off pulling image "registry.example.com/foo/bar:latest"
  Warning  Failed   2m3s (x1811 over 6h55m)  kubelet  Error: ImagePullBackOff

This means that the latest version of app001 could not be found at https://registry.example.com. This URL could be a Docker imgaes registry, a Nexus images registry, or some other type of images registry. 

This is not an issue with OpenShift. It may simply be the case that an image of app001 has not been uploaded to the registry at https://registry.example.com. Or, there might be some issue with the registry. Perhaps the registry is down, or overwhelmed with requests and timing out. The issue with the registry would need to be identified and resolved.