05-05-2013 08:46 PM
Hi Guys,
I have some inquiry regarding VCS failover service group.
This example is for 3 node cluster:
There are two service groups in <clustername> cluster, app1_SG running on node_1 and app2_SG running on node_2. In any failure, these two service groups will be switching to node_3. However, both service groups could not be running on the same node, i.e. while app1_SG switched to node_3, then app2_SG could not switch to node_3 and vice versa.
Is there a configuration on how to configure this setup?
Any comments will appreciate.
Thanks.
Solved! Go to Solution.
05-06-2013 01:57 AM
This is ideal for System Limits and Service group Prerequisites - see https://www-secure.symantec.com/connect/forums/service-group-dependency#comment-7958521
offline local group dependency is probably suitable and this would requires you give one service group a preference and that group would offline the other if it were on node3.
Mike
05-06-2013 01:49 AM
See offline local group dependency in Service group dependency configurations section in VCS admin guide.
05-06-2013 01:57 AM
This is ideal for System Limits and Service group Prerequisites - see https://www-secure.symantec.com/connect/forums/service-group-dependency#comment-7958521
offline local group dependency is probably suitable and this would requires you give one service group a preference and that group would offline the other if it were on node3.
Mike
05-06-2013 02:06 AM
Agree with Mike.
05-06-2013 06:22 AM
Just to show a difference, in Windows we would use workload managment by setting system "Limits" and service group "Prerequisites". We've used the exclusivily for Active/Actice Exhange clusters were it is impossible to run 2 virtual Exchange servers on the same physical node without causing data corruption.
I'm sure the same approach will work for other platforms as well.
-Wally