Comment 15 for bug 1928690

I'm still missing something. On the build server, I added client.nova to /etc/kolla/config/nova/ceph.client.cinder.keyring before the upgrade, and I see the client.nova lines in the container after the upgrade, but nova still can't connect:

(nova-compute)[root@chrnc-void-testupgrade-compute-0-replace /]# cat /etc/ceph/ceph.client.cinder.keyring
[client.cinder]
        key = AQCnarZgAAAAABAA/fprRD3z8dRzTgi7jtDeYA==
        caps mon = "profile rbd"
        caps osd = "profile rbd pool=volumes, profile rbd pool=vms, profile rbd pool=images"
[client.nova]
        key = AQCqarZgAAAAABAAt0lhY7TXttXIk2Y6HYQxEw==
        caps mon = "profile rbd"
        caps osd = "profile rbd pool=vms"
(nova-compute)[root@chrnc-void-testupgrade-compute-0-replace /]# ceph --id nova health
[errno 13] error connecting to the cluster
(nova-compute)[root@chrnc-void-testupgrade-compute-0-replace /]# ceph --id cinder health
HEALTH_OK

If I manually copy the key, then the problem is fixed:
(nova-compute)[root@chrnc-void-testupgrade-compute-0-replace /]# cat /etc/ceph/ceph.client.nova.keyring
[client.nova]
        key = AQCqarZgAAAAABAAt0lhY7TXttXIk2Y6HYQxEw==
        caps mon = "profile rbd"
        caps osd = "profile rbd pool=vms"
(nova-compute)[root@chrnc-void-testupgrade-compute-0-replace /]# ceph --id nova health
HEALTH_OK