On Mon, Jul 30, 2018 at 4:25 PM Dan van der Ster <dvanders@xxxxxxxxx> wrote:
>
> Hi Zheng,
>
> A stopping mds can take longer than the beacon interval.
> Before updating a cluster from 12.2.5 to 12.2.7 today I deactivated
> rank 1, and while that mds was stopping the mon thought it was *not
> responding* then replaced with a spare. So I had to again deactivate
> the new rank 1 mds.
>
> Is this expected? Shouldn't the mds continue sending the beacons while stopping?
Yes, it should be sending beacons. I'd check for other issues: was
the node overloaded, did the MDS perhaps get hung in some other way,
etc.
John
>
> 2018-07-30 10:57:44.162695 ...: audit [INF] from='client.? ...
> entity='client.admin' cmd=[{"prefix": "mds deactivate", "who":
> "cephdwightmds2"}]: dispatch
> ...
> 2018-07-30 10:59:59.489756 ... mon.0 ... : cluster [WRN] daemon
> mds.cephdwightmds2 is not responding, replacing it as rank 1 with
> standby daemon mds.cephdwightmds1
>
> Cheers, Dan
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
