yes, that's good enough for "upmap".

Mapping client features to versions is somewhat unreliable by design: not
every new release adds a new feature, some features are backported to older
releases, kernel clients are a completely independent implementation not
directly mapable to a Ceph release.

Paul

-- 
Paul Emmerich

Looking for help with your Ceph cluster? Contact us at https://croit.io

croit GmbH
Freseniusstr. 31h
81247 München
www.croit.io
Tel: +49 89 1896585 90


On Mon, Jul 29, 2019 at 2:23 PM Massimo Sgaravatto <
massimo.sgarava...@gmail.com> wrote:

> I have a ceph cluster where mon, osd and mgr are running ceph luminous
>
> If I try running ceph features [*], I see that clients are grouped in 2
> sets:
>
> - the first one appears using luminous with features 0x3ffddff8eea4fffb
> - the second one appears using luminous too, but with
> features 0x3ffddff8eeacfffb
>
> If I try to check which are these clients (I use 'ceph daemon mon.xyz
> sessions' on the 3 mons) I see that the second group includes also some
> Openstack nodes which are actually using Nautilus.
> Is this normal/expected that they appear using luminous as release ?
>
>
> Can the 'ceph features' output be used to understand if I am ready to
> switch to upmap for the balancer ?
> I.e. are 0x3ffddff8eea4fffb and 0x3ffddff8eeacfffb good enough for upmap ?
>
> Thanks, Massimo
>
>
>
> # ceph features
> {
>     "mon": {
>         "group": {
>             "features": "0x3ffddff8eeacfffb",
>             "release": "luminous",
>             "num": 3
>         }
>     },
>     "osd": {
>         "group": {
>             "features": "0x3ffddff8eeacfffb",
>             "release": "luminous",
>             "num": 70
>         }
>     },
>     "client": {
>         "group": {
>             "features": "0x3ffddff8eea4fffb",
>             "release": "luminous",
>             "num": 16
>         },
>         "group": {
>             "features": "0x3ffddff8eeacfffb",
>             "release": "luminous",
>             "num": 74
>         }
>     }
> }
> _______________________________________________
> ceph-users mailing list
> ceph-users@lists.ceph.com
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>
_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to