Stop new users from accessing the hub (I suggest by doing a k -n <hub-name> delete svc proxy-public)
Do one final rclone
Deploy the change to point to the new filestore IP. This should also bring back proxy-public svc and restore access.
(4) further needs:
k delete pv <hub-name>-home-nfs --wait=false
k -n <hub-name> delete pvc home-nfs --wait=false
k -n <hub-name> delete pod -l component=shared-dirsize-metrics
k -n <hub-name> delete pod -l component=shared-volume-metrics
This is because PersistentVolumes are not editable, and we must delete and recreate them to allow the new deploy with new IP to succeed. And this is the sequence of the objects dependent on the PV, and we need to delete all these for it to finish.
Once https://github.com/2i2c-org/infrastructure/issues/4337 is done, time to move the rest of the hubs. Should follow the same process that is tested on the staging hub.
The process here would be:
k -n <hub-name> delete svc proxy-public
)rclone
(4) further needs:
k delete pv <hub-name>-home-nfs --wait=false
k -n <hub-name> delete pvc home-nfs --wait=false
k -n <hub-name> delete pod -l component=shared-dirsize-metrics
k -n <hub-name> delete pod -l component=shared-volume-metrics
This is because PersistentVolumes are not editable, and we must delete and recreate them to allow the new deploy with new IP to succeed. And this is the sequence of the objects dependent on the PV, and we need to delete all these for it to finish.