ui: ceph: only warn about health checks that block a restart

The dialog listed every firing check and warned that a restart may be
refused. On a cluster upgraded to Ceph 19.2.6 or 20.2.4 it showed six
cephx errors and the restart then succeeded, which makes the warning
useless. Use the flag from the status endpoint, so only checks that
really block are listed.

That also needs a Force control here, which this dialog never had even
though the endpoint accepts it. A failed health query means one of two
things. A 403 only means that this user may not read the health while
the backend still can, so force stays available. Any other failure
means the backend runs into the same problem.

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>