[Pacemaker] configuration problem: migration no working always

Dejan Muhamedagic dejanmm at fastmail.fm
Wed Oct 6 15:15:56 UTC 2010


Hi,

On Wed, Oct 06, 2010 at 01:11:20PM +0200, Claus Denk wrote:
> Dear all,
> 
> I don't understand the following behaviour. We have and
> active/passive cluster with drbd, mysql and apache, I attach the
> configuration.
> I have devided the resources in two groups (filesystem+mysql in one
> "BasicServices", cluster ip and apache in the other,
> "ClusterServices"). Now, if I do a "standby" in the active node,
> everything goes nicely over to the other. Also, if I "migrate
> BasicServices", all resources migrate over to the other node.
> However, if I do "migrate ClusterServices" the resources belonging
> to the two groups are just shut down on the active node, no
> migration. Is my configuration missing something?
> 
> Thanx for any hints!!
> 
> node nodo1 \
>     attributes standby="off"
> node nodo2 \
>     attributes standby="off"
> primitive ClusterIP ocf:heartbeat:IPaddr2 \
>     params ip="192.168.1.190" cidr_netmask="32" \
>     op monitor interval="10s"
> primitive MailAlert ocf:heartbeat:MailTo \
>     params email="denk at us.es"
> primitive MySql ocf:heartbeat:mysql \
>     params binary="/usr/bin/mysqld_safe" config="/etc/my.cnf"
> socket="/var/lib/mysql/mysql.sock" datadir="/var/www/data"
> pid="/var/lib/mysql/mysql.pid" test_user="root" test_passwd=""
> test_table="claus.conceptos" \
>     op monitor interval="60s" timeout="5s" OCF_CHECK_LEVEL="1" \
>     op start interval="0s" timeout="120s" \
>     op stop interval="0s" timeout="120s"
> primitive WebFS ocf:heartbeat:Filesystem \
>     params device="/dev/drbd1" directory="/var/www" fstype="ext3"
> primitive WebSite ocf:heartbeat:apache \
>     params configfile="/etc/httpd/conf/httpd.conf" \
>     op monitor interval="1min"
> primitive nodo1-stonith stonith:external/claus \
>     params hostname="nodo1" \
>     meta target-role="Started"
> primitive nodo2-stonith stonith:external/claus \
>     params hostname="nodo2" \
>     meta target-role="Started"
> primitive ping ocf:pacemaker:ping \
>     params host_list="192.168.1.1" multiplier="100"
> primitive wwwdrbd ocf:linbit:drbd \
>     params drbd_resource="wwwdata" \
>     op monitor interval="60s"
> group BasicServices WebFS MySql \
>     meta target-role="Started"
> group ClusterServices ClusterIP WebSite MailAlert \
>     meta target-role="Started"
> ms WebData wwwdrbd \
>     meta master-max="1" master-node-max="1" clone-max="2"
> clone-node-max="1" notify="true"
> clone pingclone ping \
>     meta globally-unique="false" target-role="Started"
> location ClusterServices-on-connected-node ClusterServices \
>     rule $id="GroupServer-on-connected-node-rule" -inf: not_defined
> pingd or pingd number:lte 0
> location stonith-on-other1 nodo1-stonith -inf: nodo1
> location stonith-on-other2 nodo2-stonith -inf: nodo2
> colocation BasicServices-on-drbd inf: BasicServices WebData:Master
> colocation ClusterServices-on-BasicServcies inf: BasicServices ClusterServices

Hard to say without looking at the logs, but...

This should be the other way around:

colocation ClusterServices-on-BasicServcies inf: ClusterServices BasicServices

Perhaps that would help.

> order ClusterServices-after-BasicServices inf: BasicServices ClusterServices
> order FileSystem-after-WebData inf: WebData:promote BasicServices:start
> property $id="cib-bootstrap-options" \
>     dc-version="1.0.9-89bd754939df5150de7cd76835f98fe90851b677" \
>     cluster-infrastructure="openais" \
>     stonith-enabled="false" \
>     no-quorum-policy="ignore" \
>     expected-quorum-votes="4" \

This should be "2". Not that it really matters though.

Thanks,

Dejan

>     last-lrm-refresh="1286360161"
> rsc_defaults $id="rsc-options" \
>     resource-stickiness="100"

> 
> _______________________________________________
> 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker




More information about the Pacemaker mailing list