[Linux-cluster] How to configure resource dependency in service
Vlad Seagal
vseagal at liquidnet.com
Thu Oct 25 18:35:06 UTC 2007
Lon,
Thanks for your reply. I did the rg_test and that worked. I was able to
start the service:
[root at va0pmlt01 ~]# rg_test test /etc/cluster/cluster.conf start service
webserver
Running in test mode.
Starting webserver...
<debug> Link for bond0: Detected
<info> Adding IPv4 address 10.17.123.226 to bond0
<debug> Sending gratuitous ARP: 10.17.123.226 00:1c:c4:41:a4:1c brd
ff:ff:ff:ff:ff:ff
<info> Executing /etc/init.d/httpd start
Starting httpd: [ OK ]
Start of webserver complete
[root at va0pmlt01 ~]# ps -ef|grep http
root 28637 1 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28638 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28639 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28641 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28642 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28644 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28645 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28646 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
apache 28647 28637 0 14:23 ? 00:00:00 /usr/sbin/httpd
root 28661 8355 0 14:23 pts/0 00:00:00 grep http
[root at va0pmlt01 ~]#
[root at va0pmlt01 ~]# rg_test test /etc/cluster/cluster.conf stop service
webserver
Running in test mode.
Stopping webserver...
<info> Executing /etc/init.d/httpd stop
Stopping httpd: [ OK ]
<info> Removing IPv4 address 10.17.123.226 from bond0
Stop of webserver complete
[root at va0pmlt01 ~]#
[root at va0pmlt01 ~]# clustat
Member Status: Quorate
Member Name Status
------ ---- ------
va0pmlt01 Online, Local, rgmanager
va0pmlt02 Online, rgmanager
va0pmlt03 Online, rgmanager
va0pmlt04 Online, rgmanager
va0pmlt05 Online, rgmanager
Service Name Owner (Last) State
------- ---- ----- ------ -----
webserver (va0pmlt01) disabled
[root at va0pmlt01 ~]# clusvcadm -e webserver -m va0pmlt01
Member va0pmlt01 trying to enable webserver...success
Service webserver is now running on va0pmlt01
[root at va0pmlt01 ~]#
[root at va0pmlt01 ~]# ps -ef|grep http
root 30227 8355 0 14:32 pts/0 00:00:00 grep http
[root at va0pmlt01 ~]#
[root at va0pmlt01 ~]# clustat
Member Status: Quorate
Member Name Status
------ ---- ------
va0pmlt01 Online, Local, rgmanager
va0pmlt02 Online, rgmanager
va0pmlt03 Online, rgmanager
va0pmlt04 Online, rgmanager
va0pmlt05 Online, rgmanager
Service Name Owner (Last) State
------- ---- ----- ------ -----
webserver va0pmlt01 started
As you can see the service doesn't start properly if it's started via
clusvcadm command. Http is not running. Do you know what might be the
problem.
Thanks,
Vlad.
On 10/25/07 1:47 PM, "Lon Hohberger" <lhh at redhat.com> wrote:
> On Wed, 2007-10-24 at 15:21 -0400, Vlad Seagal wrote:
>> As you see the resource webserver doesn't get started at all.
>>
>> Here is the snap from the /etc/cluster/cluster.conf:
>>
>> <service autostart="1" domain="node1-fail" exclusive="0" name="webserver"
>> recovery="restart">
>> <clusterfs ref="GFS">
>> <ip ref="10.17.123.226">
>> <script ref="webserver"/>
>> </ip>
>> </clusterfs>
>> </service>
>
> That should work as well. If it doesn't, try running:
>
> clusvcadm -d webserver
> rg_test test /etc/cluster/cluster.conf start service webserver
>
>
> (before re-enabling, run:)
> rg_test test /etc/cluster/cluster.conf stop service webserver
>
> -- Lon
>
>
> --
> Linux-cluster mailing list
> Linux-cluster at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-cluster
>
--
Vlad Seagal
Unix System Administrator
Liquidnet Holdings, Inc.
vseagal at liquidnet.com
T 646.674.2268
More information about the Linux-cluster
mailing list