Use kubectl rollout restart instead of pod delete

[krottiers]

The chart schedule-pod-restart currently uses pod delete to trigger a restart of a specific pod. A more clean way would be to use:
'kubectl rollout restart `

https://kubernetes.io/docs/reference/generated/kubectl/kubectl-commands#-em-restart-em-

In my opinion this is a cleaner option then forcing kubernetes to discover that something is wrong and bringing everything up to spec again.