I am using 0.8.0 version of helm chart. I have set the injector failure policy to fail in the web hook. The injector and application pods are running in the same namespace. When I delete all the pods, the injector pod is not coming up with below error. Is there a way to fix this deadlock situation? This is happening intermittently. Do I need to run vault in a separate namespace?
Warning FailedCreate 3m56s (x15 over 175m) replicaset-controller Error creating: Internal error occurred: failed calling webhook “vault.hashicorp.com”: Post https://test-vault-agent-injector-svc.test.svc:443/mutate?timeout=30s: dial tcp 10.43.158.237:443: connect: connection refused