[sheepdog-users] restart cluster after update

Liu Yuan namei.unix at gmail.com
Fri May 3 12:02:30 CEST 2013


On 05/03/2013 05:37 PM, Valerio Pachera wrote:
> Hi, finally I have some time for testing.
> 
> I've been upgrading my hosts to 0.5.5_318_gc17f5d7.
> My nodes are actualy using a single disk, so I'm trying to start the
> cluster as usual.
> 
> sheep /mnt/sheepdog
> 
> Running node list I was getting
>   "There are no active sheep daemons"
> 
> So I killed alle sheep daemon and restarted corosync.
> After that, 3 nodes are showing

I think it is the corosync's problem. We don't change lines related to
corosync for ages. It looks that 47 node isn't the in ring of the
corosyncs of the cluster.

Thanks,
Yuan


More information about the sheepdog-users mailing list