[PVE-User] Proxmox 5.2, CEPH 12.2.12: still CephFS looks like jewel

Mark Schouten mark at tuxis.nl
Mon May 6 11:33:45 CEST 2019

On Mon, May 06, 2019 at 03:02:36PM +0700, Igor Podlesny wrote:
> On Mon, 6 May 2019 at 14:33, Mark Schouten <mark at tuxis.nl> wrote:
> > `ceph features` will list all connected daemons and clients.
> > Probably, there is still a client connected that runs a Jewel version.
> Can't be anything "outside" -- it's Proxmox 5.2 self-contained CEPH
> install only. No Jewel hence.

I'm not saying 'outside', I'm saying 'a client'.

> > Maybe a Kernel CephFS mount somewhere that needs a reboot?
> As I described in the very fist message this warning occurs when I
> enable CephFS on a single node of that cluster.

This command is cluster-wide, the node should not matter. But more
importantly, does `ceph features` confirm that there are no jewel
clients connected?

[Also, for someone looking for help, your responses seem somewhat rude]

Mark Schouten  | Tuxis B.V.
KvK: 74698818  | http://www.tuxis.nl/
T: +31 318 200208 | info at tuxis.nl

More information about the pve-user mailing list