virtual kubelet didn't retry for failed and pending Pods

[feiskyer]

If a Pod failed to provision due to some reasons, it would stuck in failed or pending state even after restarting virtual kubelet Pod. This should be a bug of virtual kubelet, it should retry for failed and pending Pods.

Related logs:

time="2021-02-23T08:10:32Z" level=warning msg="skipping sync of pod \"ns/pod-name\" in \"Failed\" phase" key=ns/pod-name method=syncPodInProvider name=pod-name namespace=ns node=virtual-node-aci-linux operatingSystem=Linux phase=Failed provider=azure reason= uid=18568759-17ed-492e-b131-a5e5607173a7 watchedNamespace= workerId=0

time="2021-02-23T08:10:33Z" level=info msg="Updated pod in provider" key=ns/pod-name method=createOrUpdatePod name=pod-name namespace=ns node=virtual-node-aci-linux operatingSystem=Linux phase=Pending pod=pod-name provider=azure reason=ProviderFailed uid=73d9a786-cc0c-4d40-974f-d459bea4b35f watchedNamespace= workerId=1