[Sheepdog] nodes not joing the cluster automaticaly

MORITA Kazutaka morita.kazutaka at lab.ntt.co.jp
Tue Mar 8 19:30:38 CET 2011


At Tue, 8 Mar 2011 14:02:48 +0100,
Sonali Jadhav wrote:
> We have set up sheepdog cluster for 2 KVM boxes. I have successfully installed sheepdog on both boxes. But when I try to retrieve cluster info, both boxes are showing only itself information.
> 
> On KVMbox1 its showing as,
> 
> [root at kvmbox1 ~]# collie cluster info
> running
> Ctime                Epoch Nodes
> 2011-03-08 13:04:41      1 [192.168.1.162:7000]
> [root at kvmbox1 ~]# collie node list
>   Idx        Node id (FNV-1a) - Host:Port
> ------------------------------------------------
> * 0          db28f1360c4eded6 -192.168.1.162:7000
> 
> 
> And on  KVMbox2 its showing as,
> 
> [root at kvmbox2 ~]# collie cluster info
> running
> Ctime                Epoch Nodes
> 2011-03-08 13:05:28      1 [192.168.1.163:7000]
> [root at kvmbox2 ~]# collie node list
>   Idx        Node id (FNV-1a) - Host:Port
> ------------------------------------------------
> * 0          363e4ccce90dc0c3 - 192.168.1.163:7000
> 
> I want to know why both of them are not getting added in cluster automatically? Also my corosync.conf files are configured correctly. I have configured them same as shown on http://wiki.qemu.org/Features/Sheepdog/Corosync_Config .
>  For more info I have attached sheep.log files of both boxes.
> 

It seems that your local file system doesn't support an extended
attribute.  In case of ext3, you need to add 'user_xattr' to the mount
options.

Anyway, we should print a more user friendly log message, sorry.


Thanks,

Kazutaka





More information about the sheepdog mailing list