[Gluster-users] [ovirt-users] glusterfs tips/questions

[Vijay Bellur]

On 05/23/2014 05:25 PM, Gabi C wrote:
> On problematic node:
>
> [root at virtual5 ~]# ls -la /var/lib/glusterd/peers/
> total 20
> drwxr-xr-x. 2 root root 4096 May 21 16:33 .
> drwxr-xr-x. 9 root root 4096 May 21 16:33 ..
> -rw-------. 1 root root   73 May 21 16:33
> 85c2a08c-a955-47cc-a924-cf66c6814654
> -rw-------. 1 root root   73 May 21 16:33
> c22e41b8-2818-4a96-a6df-a237517836d6
> -rw-------. 1 root root   73 May 21 16:33
> d95558a0-a306-4812-aec2-a361a9ddde3e
> [root at virtual5 ~]# cat
> /var/lib/glusterd/peers/85c2a08c-a955-47cc-a924-cf66c6814654
> uuid=85c2a08c-a955-47cc-a924-cf66c6814654
> state=3
> hostname1=10.125.1.194
> [root at virtual5 ~]# cat
> /var/lib/glusterd/peers/c22e41b8-2818-4a96-a6df-a237517836d6
> uuid=c22e41b8-2818-4a96-a6df-a237517836d6
> state=3
> hostname1=10.125.1.196
> [root at virtual5 ~]# cat
> /var/lib/glusterd/peers/d95558a0-a306-4812-aec2-a361a9ddde3e
> uuid=85c2a08c-a955-47cc-a924-cf66c6814654
> state=3
> hostname1=10.125.1.194
>

Looks like this is stale information for 10.125.1.194 that has somehow 
persisted. Deleting this file and then restarting glusterd on this node 
should lead to a consistent state for the peers.

Regards,
Vijay