Keyring problem for external rbds reported by pve8to9 check script

[Rainer Krienke]

Hello,

sorry, I need some more lines to describe the problem :-;

I have a "PveCeph" cluster and there is a second PVE cluster "PVEnoceph" 
which is using external rbds located on "PveCeph" for vm storage. So all 
VMs (~80) are running on "PVEnoceph" cluster because it has more RAM and 
CPU Power but storage is located and provided by the cluster PveCeph.

These two clusters are running fine with pve8.4, no problems. However 
running the pve9 check script /usr/bin/pve8to9 it shows an error:

     FAIL: The Ceph configuration of the following externally managed RBD
           storages needs to be updated:
           rbd_1 and rbd_2

It also offers a script to fix this. But I am reluctant, since I would 
like to know first what is wrong now and how it can be fixed without 
breaking the running clusters?

The current external rbd-configuration on "PVEnoceph" is as follows:
In PVEnoceph:/etc/pve/storage.cfg there are two rbd declarations:

rbd: rbd_1
         content images
         data-pool PveCephPool-data
         krbd 0
         monhost <ip_list of monitors for PveCeph>
         pool PveCephPool-metadata
         username admin
rbd: rbd_2
...

And in PVEnoceph:/etc/pve/priv/ceph/ there are two keyring files named
rbd_1.keyring and rbd_2.keyring which currently contain the client.admin 
Keyring of the PveCeph cluster:

[client.admin]
         key = 4711abcdef12345ABCEFxy

As said this config is running since pve 6.x without any trouble, but 
pve9 does not like the external rbds rbd_1 and rbd_2 and I have no idea why?

Any ideas?
Thanks Rainer