vinkamath

An operator error occurs roughly once a day on our H100 on which time-slicing is enabled on the `mig-1g.10gb` instances. This causes the other pods to restart as seen below...