Closed tumido closed 3 years ago
I think we need to be able to consume services over which we don't have direct administrative control. We're going to have the same issues if we consume ceph storage from NESE or from the MOC production ceph cluster after moving to new hardware.
(This also turns out to be a non-issue for operate-first/support#757, which is exclusively an OCP configuration issue).
bump
Solved - NFS integrated as an external service, we don't need to manage everything.
We need to be able to share access to the NFS cluster for operate first admins and oncalls - to be able to debug possible issues like: https://github.com/operate-first/SRE/issues/333
Possibly we should thing about making this cluster part of Operate First and make it's setup declarative as well.
/cc @HumairAK @larsks @4n4nd thoughts, ideas how to do that?