I have a project in west-1 on Openshift v3. In it I have an app that worked fine and one that stopped working following a GitHub committing something very downstream within the code. The issue is with making a pod:
No nodes are available that match all of the following predicates::
Insufficient cpu (173), MatchNodeSelector (5). I stopped it and then I scaled down the other pod to see if it was any weird settings that the new build for the other might have picked up and I get the same. Now both are broken.
Is it that Red Hat Openshift has run out of CPUs or have I got some weird default setting that has decided to request 173 CPUs and 5 Nodes?
The message doesn't mean that the pod is requesting 173 CPUs.
The error message means that scheduling failed that on 173 machines there isn't enough CPU left available to fit the pod requests (see pod definition) and on 5 machines (nodes) it can't be scheduled because those nodes don't have the particular set of labels the pod is specifying in spec.nodeSelector of the pod definition. (There might be intersection between the 2 groups of nodes)