[Pacemaker] Problems cleaning up resource with hard errors

[Rafael Porres Molina]

Finally I was able to spot the problem.

It turns out that the resource was not properly configured. Somehow /etc/init.d/mysqld was working properly but resource.d/heartbeat/mysql was not. Even if I pointed both to the same configuration file, the script was ignoring some of the parameters I had set in the config file. It was quite useful to run directly the resource script setting the OCF variables as shown:

OCF_ROOT=/usr/lib/ocf
OCF_RESKEY_test_user=p_mon
OCF_RESKEY_config=/etc/my.cnf
OCF_RESKEY_test_table=monitor.boolean
OCF_RESKEY_group=mysql
OCF_RESKEY_test_passwd=p_mon
OCF_RESKEY_user=mysql
OCF_RESKEY_socket=/var/lib/mysql/mysql.sock
OCF_RESKEY_datadir=/opt/manhattan/mysql
OCF_RESKEY_binary=/usr/bin/mysqld_safe

The configuration for the mysql resource is now:

primitive p_mysqld_left ocf:heartbeat:mysql \
	params binary="/usr/bin/mysqld_safe" config="/etc/my.cnf" test_table="monitor.boolean" test_user="p_mon" test_passwd="p_mon" datadir="/opt/manhattan/mysql" socket="/var/lib/mysql/mysql.sock" user="mysql" group="mysql" \
	op monitor interval="15s" timeout="30s" \
	op start interval="0" timeout="120s" \
	op stop interval="0" timeout="120s" \
	meta target-role="Started"

Best regards,

Rafa

-- 
Rafael Porres Molina
Consultoría y Proyectos
Qindel Formación y Servicios S.L.

Móvil: (+34) 678650609
e-mail: rafael.porres at qindel.com
Dirección: c/Julián Camarillo 29, Edificio D2, 4ºIzda, 28037 Madrid, SPAIN / ESPAÑA






-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20100728/4aaf4540/attachment-0001.htm>