amin's questions

I have an active/passive high availability cluster with pacemaker+split brain. After split brain condition I use automatic split-brain recovery to recover my systems.

The problem is that since file system resource is managed by CRM it is not mounted after split brain recovery. Nothing is mounted after split brain recovery!

It seems I should use stonith devices, but my boss ask me to solve it without any extra device! How can I notify the split brain situation to pacemaker so it will remount the file system resource?

EDIT: my crm config

node drbd3
node drbd4
primitive apache ocf:heartbeat:apache \
        params configfile="/etc/apache2/apache2.conf" httpd="/usr/sbin/apache2" \
        op monitor interval="5s"
primitive drbd_disk ocf:linbit:drbd \
        params drbd_resource="r0" \
        op monitor interval="15s"
primitive fs_drbd ocf:heartbeat:Filesystem \
        params device="/dev/drbd/by-res/r0" directory="/mnt" fstype="ext4"
primitive ip1 ocf:heartbeat:IPaddr2 \
        params ip="10.8.0.95" nic="eth0"
group group1 ip1 apache
ms ms_drbd drbd_disk \
        meta master-max="1" master-node-max="1" clone-max="2" clone-node-max="1" notify="true"
colocation mnt_on_master inf: fs_drbd ms_drbd:Master
order apache_after_ip inf: ip1:start apache:start
order mount_after_drbd inf: ms_drbd:promote fs_drbd:start
property $id="cib-bootstrap-options" \
        dc-version="1.0.8-042548a451fce8400660f6031f4da6f0223dd5dd" \
        cluster-infrastructure="openais" \
        expected-quorum-votes="2" \
        stonith-enabled="false" \
        no-quorum-policy="ignore"

and my drdb config:

global { usage-count no; }
common { syncer { rate 100M; } }
resource r0 {
        protocol C;
        startup {
                wfc-timeout  1;
                degr-wfc-timeout 1;
        }
#       disk{
#               fencing resource-only;
#       }
        handlers {
                #split-brain "/usr/lib/drbd/notify-split-brain.sh root";
#                fence-peer "/usr/lib/drbd/crm-fence-peer.sh";
#                after-resync-target "/usr/lib/drbd/crm-unfence-peer.sh";

        }

        net {
                cram-hmac-alg sha1;
                shared-secret "test";
               # allow-two-primaries;
                after-sb-0pri discard-zero-changes;
                after-sb-1pri discard-secondary;
                after-sb-2pri call-pri-lost-after-sb;
        }
        on drbd3 {
                device /dev/drbd0;
                disk /dev/sda1;
                address 10.8.0.93:7788;
                meta-disk internal;
        }
        on drbd4 {
                device /dev/drbd0;
                disk /dev/sda1;
                address 10.8.0.94:7788;
                meta-disk internal;
        }
}
#include "drbd.d/global_common.conf";
#include "drbd.d/*.res";
~                          

I am trying to avoid a split brain situation using automatic split brain recovery policies explained at http://www.drbd.org/users-guide/s-configure-split-brain-behavior.html

How to predict which condition we are involved in? How to predict the number of primaries after split brain? Is it all accidental?

i am planning for high availability on lucid ubuntu, i started with drbd after a simple configuration on drbd I went to heartbeat and had a simple config on it too.

on drbd official site a document told me heartbeat is legacy and i should meet pacemaker instead as CRM.

on pacemaker related documents i understand heartbeat and OpenAis both are cluster infrastructures but openais is more preferred choosed by redhat too, so i went to openais and installed openais after searching with openais i understand that corosync is a project come from openais and it seem that corosync is used with ubuntu instead of openais.

and i want to configure corosync but no config is present where document declare!

1.could any body describe role of pacemaker in relation with openais,corosync ?

2.what's exactly preferred technology for high availability on ubuntu?

3. could you please give a step by step guide to set high availability on ubuntu?

thanks in advance.

some filesystems like btrfs and lustre are marked as fault tolerant file systems in : wiki:http://en.wikipedia.org/wiki/List_of_file_systems

what's meaning of fault tolerant file system at all? does it offer some thing more than journaling?

if you enable block journaling in ext3 you can call it fault tolerant or these file systems offer some thing more? my field of interest is to know what these file systems offer about power failure problems?

thanks