CMO condition messages are not very helpful when the managed objects (such as deployments) are unavailable/degraded.

   At least 4.14 

    Always in case of error

    1. Configure a node selector for 1 monitoring component which fails the scheduling (e.g. "nodeSelector: {foo: bar}") 
    2. Wait for a couple of minutes till CMO reports Available=False & Degraded=True conditions
    3.
    

    the Available condition's message is unhelpful: "UpdatingTelemeterClient: reconciling Telemeter client Deployment failed: updating Deployment object failed: waiting for DeploymentRollout of openshift-monitoring/telemeter-client: context deadline exceeded"

An indication about the telemetry-client deployment state.