[Pacemaker] How to disallow resources to re-run after they were stopped

Andrew Beekhof andrew at beekhof.net
Mon Oct 7 23:13:08 UTC 2013


On 01/10/2013, at 8:09 PM, Mistina Michal <Michal.Mistina at virte.sk> wrote:

> Dear all.
> I am using ping resource agent to check if the resources can run on the node. I have 2 node cluster.
> If the destination is down resources are also down on that particular node. If the destination can be pinged again the resources boot again on the node.
> My question is, how can I achieve the node which once couldn’t ping the destination is marked as the host where the resources cannot run anymore, so the resources stay in stopped state?

You would have to modify the agent to create a location constraint with score=-INFINITY when the destination because unreachable.
The better question though, is why you would want this.

>  
> Versions:
> OS: RHEL 6.3
> pacemaker-1.1.7-6
> corosync-1.4.1-7
> resource-agents-3.9.2-12
>  
> Config:
> node node1
> node node2
> primitive drbd_pg ocf:linbit:drbd \
>         params drbd_resource="postgres" \
>         op monitor interval="15" role="Master" \
>         op monitor interval="16" role="Slave" \
>         op start interval="0" timeout="240" \
>         op stop interval="0" timeout="120"
> primitive pg_fs ocf:heartbeat:Filesystem \
>         params device="/dev/vg_local-lv_pgsql/lv_pgsql" directory="/var/lib/pgsql/9.2/data" options="noatime,nodiratime" fstype="xfs" \
>         op start interval="0" timeout="60" \
>         op stop interval="0" timeout="120"
> primitive pg_lsb lsb:postgresql-9.2 \
>         op monitor interval="30" timeout="60" \
>         op start interval="0" timeout="60" \
>         op stop interval="0" timeout="60" \
>         meta is-managed="true"
> primitive pg_lvm ocf:heartbeat:LVM \
>         params volgrpname="vg_local-lv_pgsql" \
>         op start interval="0" timeout="30" \
>         op stop interval="0" timeout="30"
> primitive pg_ping ocf:pacemaker:ping \
>         params dampen="5s" host_list="192.168.100.254" multiplier="101" attempts="10" \
>         op monitor interval="15s" timeout="30s"
> primitive pg_pool ocf:heartbeat:pgpool \
>         params pgpoolconf="/etc/pgpool-II/pgpool.conf" pcpconf="/etc/pgpool-II/pcp.conf" hbaconf="/etc/pgpool-II/pool_hba.conf" logfile="/var/log/pgpool/pgpool.log" pgpooluser="postgres" checkmethod="pid" checkstring="/var/run/pgpool/pgpool.pid" pgpoolcmd="/usr/bin/pgpool" psqlcmd="/usr/pgsql-9.2/bin/psql" pcpnccmd="/usr/bin/pcp_node_count" \
>         op monitor interval="20" timeout="30" \
>         op start interval="0" timeout="20" \
>         op stop interval="0" timeout="20" \
>         meta is-managed="true"
> primitive pg_vip ocf:heartbeat:IPaddr2 \
>         params ip="192.168.100.20" iflabel="nodevip" \
>         op monitor interval="5"
> group PGServer pg_lvm pg_fs pg_vip pg_lsb pg_pool \
>         meta target-role="Started"
> ms ms_drbd_pg drbd_pg \
>         meta master-max="1" master-node-max="1" clone-max="2" clone-node-max="1" notify="true" target-role="Started"
> clone clone_pg_ping pg_ping \
>         meta target-role="Started"
> location master-prefer-node1 pg_vip 50: node1
> location promote_master_on_node_w_reachable_hosts ms_drbd_pg \
>         rule $id="promote_master_on_node_w_reachable_hosts-rule" $role="master" -inf: pingd lte 0
> location run_grp_on_node_w_reachable_hosts PGServer \
>         rule $id="run_grp_on_node_w_reachable_hosts-rule" -inf: pingd lte 0
> colocation col_pg_drbd inf: PGServer ms_drbd_pg:Master
> order ord_pg inf: ms_drbd_pg:promote PGServer:start
> property $id="cib-bootstrap-options" \
>         dc-version="1.1.7-6.el6-148fccfd5985c5590cc601123c6c16e966b85d14" \
>         cluster-infrastructure="openais" \
>         expected-quorum-votes="4" \
>         stonith-enabled="false" \
>         no-quorum-policy="ignore" \
>         maintenance-mode="false" \
>         last-lrm-refresh="1379343052"
> rsc_defaults $id="rsc-options" \
>         resource-stickiness="100"
>  
> Any help would be appreciated.
>  
> Best regards,
> Michal Mistina
>  
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 841 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20131008/3c42dd5e/attachment-0004.sig>


More information about the Pacemaker mailing list