J C Lawrence

Yes, this would be for bucket access -- same as using an instance role in AWS. (We have the same problem/concern)

Yep, making it optional would be good. GoogleManagedPrometheus borks on `partial_response` (and largely removes any need to run Thanos), so I'm currently running a rewriting proxy atop Prometheus to filter...

This weekend `gmp-operator` started crashlooping with, "no endpoints available for service gmp-operator". While there were no OOMKill events, that turned out to be the cause. Putting in a VPA addressed....

Yes, in that deploys would have worked despite the failure. Being unable to eg ship a critical security fix due to GMP indigestion is not right. Additionally, the lack of...

Supplemental in case useful -- heading into and during the failures: 

Making self-observation opt-in is sensible. Something as simple as a doc page or reference pointing at reference pod monitors and possible alert rules under https://g.co/kgs/dDdszEW would help a lot. In...