In the Firefly release of Ceph, all OSDs asked the mons every question about maps that they had. In the Jewel release, that burden was taken off of the Mons and given to the OSDs as often as possible to ask each other. This was done because the Mons were found to be a bottleneck for growing a cluster very large. I had a cluster on Firefly that the mons couldn't handle the load if the cluster grew larger than 1200 OSDs. Now in Luminous they are again removing some strain on the Mon daemons to allow clusters to continue to grow to a larger scale. Like John mentioned, this new daemon is there to remove strain on the Mons.
"It is not mandatory to place mgr daemons on the same nodes as mons, but it is almost always sensible."
This quote from John's PR for the ceph documentation shows that this is not a change for standard Ceph usage, but for extreme clusters pushing the bounds of how large you can grow a cluster. Few times will anyone need to get that much more performance from their mons, but it is awesome that they have created this new daemon to be able to isolate these tasks when you do need it for scale.
On Wed, Jun 14, 2017 at 10:29 AM John Spray <jspray@xxxxxxxxxx> wrote:
On Wed, Jun 14, 2017 at 6:47 AM, Manuel Lausch <manuel.lausch@xxxxxxxx> wrote:
> Hi,
>
> we decided to test a bit the upcoming ceph release (luminous). It seems
> that I need to install this ceph-mgr daemon as well. But I don't
> understand exactly why I need this service and what I can do with it.
>
> The ceph Cluster is working well without installing any manager daemon.
> However in the ceph status output there is a health error ("no active
> mgr")
>
> My questions:
> Can we run the cluster without this additional daemon? If yes, is it
> possible to supress this health error?
You do need the daemon -- in luminous, it is responsible for
populating parts of the health monitoring, especially the status of
PGs.
> What exactly is the purpose of this service. The documentation oncontains
> very rare information about it.
You probably already found the existing doc page for mgr:
"The Ceph Manager daemon (ceph-mgr) runs alongside monitor daemons, to
provide additional monitoring and interfaces to external monitoring
and management systems."
I've fleshed it out a little bit: https://github.com/ceph/ceph/pull/15690
John
>
>
> Regards
> Manuel
>
>
>
>
>
> --
> Manuel Lausch
>
> Systemadministrator
> Cloud Services
>
> 1&1 Mail & Media Development & Technology GmbH | Brauerstraße 48 |
> 76135 Karlsruhe | Germany Phone: +49 721 91374-1847
> E-Mail: manuel.lausch@xxxxxxxx | Web: www.1und1.de
>
> Amtsgericht Montabaur, HRB 5452
>
> Geschäftsführer: Frank Einhellinger, Thomas Ludwig, Jan Oetjen
>
>
> Member of United Internet
>
> Diese E-Mail kann vertrauliche und/oder gesetzlich geschützte
> Informationen enthalten. Wenn Sie nicht der bestimmungsgemäße Adressat
> sind oder diese E-Mail irrtümlich erhalten haben, unterrichten Sie
> bitte den Absender und vernichten Sie diese E-Mail. Anderen als dem
> bestimmungsgemäßen Adressaten ist untersagt, diese E-Mail zu speichern,
> weiterzuleiten oder ihren Inhalt auf welche Weise auch immer zu
> verwenden.
>
> This e-mail may contain confidential and/or privileged information. If
> you are not the intended recipient of this e-mail, you are hereby
> notified that saving, distribution or use of the content of this e-mail
> in any way is prohibited. If you have received this e-mail in error,
> please notify the sender and delete the e-mail.
>
> _______________________________________________
> ceph-users mailing list
> ceph-users@xxxxxxxxxxxxxx
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
