cancel
Showing results for 
Search instead for 
Did you mean: 

VCS 5.0 MP3 version is not working

Sridhar_sri
Level 5
Hi Gurus,
                
We have Two servers in VCS setup, Named A and B with our product installed. In both of our server, the corresponding service were not running.  we have written on our agent scripts (i.e. online.pl, offline.pl, monitor.p) and imoprted our product_types.cf successfully using Veritas cluster server manager java console.

This is a HA setup made up with 2 nodes. After configuring cluster server, we have created a service group called AR with type as failover. Followed by created a resource called chk_r(using our product resource) under this service group and marked both Critical and Enabled option. When we try to ONLINE this resource, it throws the following error in engine_A.log

22:36:29 VCS ERROR V-16-1-10303 Resource chk_r (Owner: unknown, Group: chk) is FAULTED (timed out) on sys odc-popc-9
22:34:25 VCS ERROR V-16-10001-102 (odc-popc-9) AR:chk_r:monitor:The AR for chk_r did not have the expected State.

After few minutes it moves Server A (where online is tried) to faulted state.

Can anyone help me to solve this problem


Thanks,
Sri.
3 REPLIES 3

mobilegleed
Level 3
Employee
Hi Sri,

I assume this is a Windows Cluster?  What commands are you running inside your monitor.pl to determine the health of the services?  From the errors it appears the monitor.pl is not returning with the expected return code.  In the case the service is running do you use the exit value of 110, and 100 when it's not running?

Have you looked at the GenericService Agent?  If the these are just Windows Services you may find it's easier to use this agent opposed to writing your own scripts.

Regards,
-Kyle

Tavitayya
Not applicable


Hi Kyle,

Thanks for your respone .

It is Solaris CLustering and as you said  the exit value for working is 110 and not working 100.

Thanks,
Tavitayya

Gaurav_S
Moderator
Moderator
   VIP    Certified
Hi,

As you say that exit codes are correct & still above messages are coming.... so I understand script is not able to correctly determine the state...

in this case:

a) Can you bring resource online manually ? also outside to VCS ?
b) Can you probe resource successfully ?
c) If you run the monitor script manully, what exit code it returns ? (echo $?)


Gaurav