[Linux-cluster] MySQL service

Marek 'marx' Grac mgrac at redhat.com
Mon Sep 22 11:58:20 UTC 2008


gregory steulet wrote:
> Hi folks,
> I've a problem with the mysql resource. I cannot start/stop or even monitor a service with mysql. You should find everything regarding my configuration and the error messages below. For Information I manage other resources such as VIP that are working fine.
> 1.Check mysql status 
> ----------------------
> [root at emperor01 ~]# /etc/init.d/mysql.server status
> MySQL is not running                                       [FAILED]
> 3.Check mysql status 
> -----------------------
> [root at emperor01 ~]# /etc/init.d/mysql.server status
> MySQL is not running                                       [FAILED]
When you use a RH Cluster Suite and our resource agent then you should 
NOT use init scripts in /etc/init.d because they are not cluster ready. 
Main advantage of resource agent is that it is possible to have several 
Mysql/Apache/... running on same machine. Standard init scripts checks 
pid, lock files etc in standard paths. We have them in different paths.

> 6. cluster.conf
> --------------------
>                 <resources>
>                         <ip address="" monitor_link="1"/>
>                         <mysql config_file="/etc/my.cnf" listen_address="" name="mysql" shutdown_wait="0"/>
>                 </resources>
>                 <service autostart="1" exclusive="0" name="mysql" recovery="relocate">
>                         <ip ref=""/>
>                 </service>
>                 <service autostart="1" name="mysql_service">
>                         <mysql ref="mysql"/>
>                 </service>
>         </rm>

> 7. /usr/share/cluster/mysql.sh
> --------------------------------
> ..export LC_ALL=C
> export LANG=C
> export PATH=/bin:/sbin:/usr/bin:/usr/sbin
>  $(dirname $0)/ocf-shellfuncs
>  $(dirname $0)/utils/config-utils.sh
>  $(dirname $0)/utils/messages.sh
>  $(dirname $0)/utils/ra-skelet.sh
> declare MYSQL_MYSQLD=/u00/app/mysql/product/5.1.26/bin/mysqld_safe
> declare MYSQL_ipAddress
> declare MYSQL_pid_file=/u00/app/mysql/admin/mysqld1/socket/mysqld1.pid
> declare MYSQL_timeout=30
> ..

In normal use there is no need to change resource agent itself. You need 
to change MYSQL_MYSQLD and this should not broke anything. From what I 
can see from your output problem is that we are waiting for pid file 
(line 126 of mysql.sh) and it appears elsewhere or didn't appear in 30 


Marek Grac
Red Hat Czech s.r.o.

More information about the Linux-cluster mailing list