Hi,

On Mon, Nov 05, 2012 at 01:20:05PM +0900, Josh Bowling wrote:
> Hi all,
> 
> I'm nearly at the end of setting up my very first
> KVM+DRBD+Pacemaker+Corosync cluster and would like for someone a bit more
> experienced to take a look at my final config and let me know if I'm doing
> anything stupid.  I would especially like some critiques on the stonith
> setup:
> 
> > node carbon
> > node oxygen
> > primitive neon_vm ocf:heartbeat:VirtualDomain \
> >         params config="/etc/libvirt/qemu/neon.xml"
> > hypervisor="qemu:///system" \
> >         meta allow-migrate="true" \
> >         op start interval="0" timeout="120s" \
> >         op stop interval="0" timeout="120s" \
> >         op monitor interval="10" timeout="30" depth="0"
> > primitive nickel_vm ocf:heartbeat:VirtualDomain \
> >         params config="/etc/libvirt/qemu/nickel.xml"
> > hypervisor="qemu:///system" \
> >         meta allow-migrate="true" \
> >         op start interval="0" timeout="120s" \
> >         op stop interval="0" timeout="120s" \
> >         op monitor interval="10" timeout="30" depth="0"
> > primitive res_drbd_disk0 ocf:linbit:drbd \
> >         params drbd_resource="disk0" \
> >         op monitor interval="120s" \
> >         op start interval="0" timeout="240s" \
> >         op stop interval="0" timeout="100s"
> > primitive res_fs ocf:heartbeat:Filesystem \
> >         params device="/dev/drbd0" directory="/shared" fstype="ext4" \
> >         meta target-role="Started"
> > primitive res_ip ocf:heartbeat:IPaddr2 \
> >         params ip="192.168.10.150" cidr_netmask="24" nic="eth0"
> > primitive silicon_vm ocf:heartbeat:VirtualDomain \
> >         params config="/etc/libvirt/qemu/silicon.xml"
> > hypervisor="qemu:///system" \
> >         meta allow-migrate="true" \
> >         op start interval="0" timeout="120s" \
> >         op stop interval="0" timeout="120s" \
> >         op monitor interval="10" timeout="30" depth="0"
> > primitive st-ipmilan stonith:ipmilan \
> >         params hostname="carbon" ipaddr="192.168.10.151" port="623"
> > auth="straight" priv="admin" login="admin" password="secret"
> > primitive st-ipmilan2 stonith:ipmilan \
> >         params hostname="oxygen" ipaddr="192.168.10.152" port="623"
> > auth="straight" priv="admin" login="admin" password="secret"

Better use external/ipmi rather than ipmilan. The use is very
similar.

You should also add the negative location preferences for the two
stonith resources:

location pref-stonith-oxygen st-ipmilan2 -inf: oxygen
location pref-stonith-carbon st-ipmilan -inf: carbon

> >         rule $id="drbd-fence-by-handler-rule-ms_drbd_disk0" $role="Master"
> > -inf: #uname ne carbon

> > primitive zinc_vm ocf:heartbeat:VirtualDomain \
> >         params config="/etc/libvirt/qemu/zinc.xml"
> > hypervisor="qemu:///system" \
> >         meta allow-migrate="true" \
> >         op start interval="0" timeout="120s" \
> >         op stop interval="0" timeout="120s" \
> >         op monitor interval="10" timeout="30" depth="0"
> > group rg_disk0 res_fs res_ip silicon_vm zinc_vm neon_vm nickel_vm
> > ms ms_drbd_disk0 res_drbd_disk0 \
> >         meta notify="true" master-max="1" master-node-max="1"
> > clone-max="2" clone-node-max="1"
> > location drbd-fence-by-handler-ms_drbd_disk0 ms_drbd_disk0 \
> >         rule $id="drbd-fence-by-handler-rule-ms_drbd_disk0" $role="Master"
> > -inf: #uname ne carbon
> > colocation c_disk0_on_drbd inf: rg_disk0 ms_drbd_disk0:Master
> > order o_drbd_before_vms inf: ms_drbd_disk0:promote rg_disk0:start
> > property $id="cib-bootstrap-options" \
> >         dc-version="1.1.6-9971ebba4494012a93c03b40a2c58ec0eb60f50c" \
> >         cluster-infrastructure="openais" \
> >         expected-quorum-votes="3" \

How many nodes do you have? Not that it matters in this case, but
still.

> >         stonith-enabled="true" \
> >         no-quorum-policy="ignore" \
> >         default-action-timeout="240" \
> >         stonith-action="poweroff"
> >
> 
> I'm planning on putting this cluster live in a couple of days, so I'd
> really like to make sure I don't do something stupid. ;)

Hope you did some thorough testing.

Thanks,

Dejan

> Thanks
> _______________________________________________
> Linux-HA mailing list
> [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to