datastax
Proxy configmap is not automatically re-created after deletion
If I manually delete the proxy configmap, the operator can get into a bad state where it's waiting for the proxy pods to come up, but they can't come up because the configmap is missing.
A workaround for this was to disable and then re-enable TLS on the proxy to force a new config to be created.
The operator doesn't watch the generated resources. In order to implement this, we have to use the JOSDK secondary resource concept. This would be a huge change and I'm not sure it's worth
It seems like the operator should have some way to bring the cluster back into a working state. Since the operator creates the configmap it should also include some way to re-sync it. Maybe have a timeout on proxy deployment readiness where it will re-apply the configmap if the proxy, or other resource, fails to become ready? Or check all resources at startup so that an operator restart will fix the issue?