Description of problem: CMO doesn't report any problems with monitoring stack when alertmanager cannot set up gossip network. Version-Release number of selected component (if applicable): How reproducible: Found it when trying to setup cluster-wide proxy Steps to Reproduce: 1. Start a cluster 2. Setup cluster-wide proxy 3. Look into alertmanager logs Actual results: Alertmanager cannot establish gossip network, but CMO is not in degraded state Expected results: CMO in degraded state due to problems with alertmanager Additional info:
Alertmanager not being able to establish a gossip network is not good, but doesn't mean a cluster is degraded. People still get notifications (even if duplicate), and running full e2e test suite on a cluster is also exaggerated. I don't think this is a bug.
Closing as NOTABUG. Please reopen if you feel further action is required.