[Linux-cluster] Monitoring services/customize failure criteria

[Chen, Mockey (NSN - CN/Cheng Du)]

>> How can I customize "failure criteria" for a cluster resource?
>
>If you are using a Script resource, the script will get called 
>with a 'status' parameter. It is totally up to you how you 
>validate the resource. Using Apache as an example, you can 
>check for a pid file, check that at least one httpd process 
>shows up in the process list, open a socket connection to the 
>proper port to make sure it's listening, send a complete HTTP 
>request to httpd and parse the output - the possibilities are 
>many. It's your decision how you want to do the validation.

In default configuration, cluster check the resource health every 30
seconds, 
I want to config this interval in /etc/cluster/cluster.conf, but I can
not find any related parameter,
or I missing something ?

The only way I can change it is mannually change
/usr/share/cluster/script.sh

Any suggestions ?


Regards.