[Cephadm][nfs-ganesha] Migration of old nfs daemons failing after dropping the NFSServiceSpec pool and namespace properties.

Issue

• Migration of the NFS services from RHCS 16.2.5 (or earlier releases) to higher versions is failing.

• Ceph Cluster is in HEALTH_ERR state with following alerts:

  cluster:
    id:     3e5649b6-93fb-11ec-9592-5c6f697228f0
    health: HEALTH_ERR
            Module 'cephadm' has failed: [errno 2] RADOS object not found (error opening pool 'nfs-ganesha')