On the first run, the istio-operator doesn't have access to some secrets, so it waits 60 seconds and then terminates with a non-zero exit code in order to be restarted by the Kubelet. The problem with this is that the Pod's status is shown as `Error`:

istio-operator-6f684485b-rdkb8     0/1     Error               0               54s
istio-operator-6f684485b-rdkb8     1/1     Running             1 (8s ago)      61s

We'll improve this so that the restart happens sooner and without showing this as an `Error`, since the restart is expected.