On 2/9/07, Brandon Lamb <brandonlamb@gmail.com>
wrote:> Hey
>
> So I installed a new setup today as follows
>
> server1
> 17G raid /dev/sda1 using drbd v8 set as /dev/drbd0
>
> server2
> 17G raid /dev/sda1 using drbd v8 set as /dev/drbd0
>
> The drbd seems to work fine so far.
>
> I also have
>
> cluster1 and cluster2 nodes.
>
> cluster1 connects to server1 using iscsi
> cluster2 connects to server2 using iscsi
>
> On cluster1 i did
> mkfs.ocfs2 -N 5 -T mail -L ocfs2 /dev/sda
>
> This seemed to go fine (ocfs2 was loaded and my ocfs2 cluster was
> online on both nodes)
>
> I was able to mount the ocfs2 drive on cluster1
>
> On cluster2 when I try to mount I get
> [root@cluster2 mnt]# mount cluster
> ocfs2_hb_ctl: OCFS2 inode is not a directory while reading uuid
> mount.ocfs2: Error when attempting to run /sbin/ocfs2_hb_ctl:
> "Operation not permitted"
>
> What did I miss? I am guessing its because of the drbd setup as this
> same config was working for just 1 server.
>
> Do I have to do something special? Do I make cluster1 its own cluster
> and cluster2 its own cluster and let drbd handle the drive replication
> or...
>
> Im confused
>
> =)
>
> Brandon
I just found and applied this patch
http://lists.linbit.com/pipermail/drbd-dev/2007-January/000613.html
I am rebooting now to see if this was maybe my problem.