Yes I have encountered some issues on ocfs2 but these were not split-brain
problems. Also my problems reasons were about the SPOF. So I want to use Drbd.
I will join to drbd maillist and ask this subject on there again.
Thanks,
Igor Cicimov wrote:
>On Wed, Dec 26, 2012 at 11:47 PM, Atıf CEYL
On Wed, Dec 26, 2012 at 11:47 PM, Atıf CEYLAN wrote:
> **
> On Wed, 2012-12-26 at 23:12 +1100, Igor Cicimov wrote:
>
> Maybe try heartbeat if tou are after something simple. Using dual primary
> though without fencing is asking for trouble, split brain and lost od data.
>
> Yes. maybe I will enc
On Wed, 2012-12-26 at 23:12 +1100, Igor Cicimov wrote:
> Maybe try heartbeat if tou are after something simple. Using dual
> primary though without fencing is asking for trouble, split brain and
> lost od data.
>
Yes. maybe I will encounter a split brain problem. I had asked the first
question f
You are saying me that use HA and you have a master file system and
share from on it via NFS to the other server. is that right?
I tried some scenarios about cluster. GlusterFS, NFS and OCFS2. My
system daily load is very highly. It's over 50 million transactions
daily. So nfs and glusterfs are not
Thanks for the answer.
if I must do it, I don't want to use the HA (pacemaker or corosync). I
should run on both server actively. There is a way for active-active
cluster without any HA solution? Do I know incorrect thi?
On Wed, 2012-12-26 at 00:43 +1100, Igor Cicimov wrote:
> Check pacemaker the
Hi all,
I have Apache and Postfix+Dovecot are running on two debian servers. I
want to use drbd+ocfs2 (or redhat gfs).
I tried some configurations and fault scenarios on VM but I have data
loss. Would you like suggest me best practice about drbd+cluster
filesystem with two-primary mode?
regard
6 matches
Mail list logo
