[Linux-HA] Monitoring resources (apache)

[Joseph Andres]

 Hi All,

I'm very new to Linux  HA, nevertheless have managed to install a 2 node
cluster in my test environment. My test environment is:

1 x Laptop running CentOS 4.4 (kernel 2.6.19.1)
1 x Laptop Windows XP plus VMWare running CentOS 4.4 (kernel 2.6.19.1)
Heartbeat v 2.0.8

I have installed HB using RPM from CentOS repository.

The cluster works fine, the failover is successful in the following
simulated scenarios:

- node crash, i.e. poweroff or shutdown a node
- making one node standby
- stopping HB in one node

The next level I'm trying to achieve is to monitor resources. However, this
is where I got stuck. I have only two resources now, IPaddr and apache. I'm
trying to monitor apache with the attached configuration in my
cib.xml(output of cibadmin -Ql command). I simpIy want to failover to
the other
node, i.e. migrate all resources, if apache has stopped for some reasons (
e.g. service httpd stop). I tried to follow the examples in
http://www.linux-ha.org/ClusterInformationBase/Actions.

With my current configurations, however, my two nodes just going in loop
failing over (bouncing) and seems like never able to start apache properly,
or may be the monitor itself is failing does causing the loop. It's really
"interesting" looking at it with hb_gui. When this happens, I put back my
original config using:

cibadmin -E
cibadmin -C -o resources -x all_resources
cibadmin -C -o constraints -x constraints

(Btw, I also had to separate my "all_resources" and "constraints" base files
as above because they cannot seem to go together. Is this normal?)

Kindly advice how could I possibly fix this problem.

Thanks.
--Joseph
-------------- next part --------------
A non-text attachment was scrubbed...
Name: monitoringresources.zip
Type: application/zip
Size: 16298 bytes
Desc: not available
Url : http://lists.community.tummy.com/pipermail/linux-ha/attachments/20070313/edd30e4a/monitoringresources.zip