Video Screencast Help

Force VCS failover for just one "hop"

Created: 05 Aug 2010 • Updated: 08 Sep 2010 | 3 comments
This issue has been solved. See solution.

Hi all,
this is our environment: a VCS 5.1 cluster on 5 nodes REHL 5.5 x86_64.
I've configured round robin as failover policy for all service group in the cluster.

Now I'm asking if there is some options or attribute for forcing the failover of a service group for just one hop.
Explaination: with round robin policy a service group failover on the system with the least number of active service groups. And that's ok. But in an application problem scenario the SG failover on every node available in the cluster before stopping. I want to keep the round robin policy but I'want that after the first failover the cluster stop migrate the failing SG.

Is it possible?

Thanks and have a nice day

Comments 3 CommentsJump to latest comment

Gaurav Sangamnerkar's picture

Hello,

I don't think that would be a possible option.... reason being, since you have populated your systemlist with all 5 nodes (which I hope), which means you have mentioned that apps can failover to any node....

Now if there is some application problem, HAD engine will decide the failover target, in your case it would be the node which is running with least service groups.... lets say if SG doesn't comes online after failover, it will continue to attempt to bring service group online on other nodes..... there are couple of SG attributes which can control restart attempts, however that is on same system, there should be a attribute like failoverattempt (which you can suggest as an idea for next release)...

As far my knowledge, it might not be possible..

Gaurav

PS: If you are happy with the answer provided, please mark the post as solution. You can do so by clicking link "Mark as Solution" below the answer provided.
 

SOLUTION
vcsuser's picture

Thanks for you answer Gaurav! I think I have to modify my needs...

Have a nice day

rregunta's picture

Hello,

VCS cluster depends upon 2 vairables - SystemList and AutoStartList for each service group start and fail over -

Eg. 


group group1 (
SystemList = { Server1, Server2 }
AutoStartList = { Server1 }

VCS uses SystemList when the cluster brings up the service group for the first time in the cluster. when the Service groups are UP and running the FailOverPolicy takes over and it refers to AutoStartList for each service group. So when any service group faults, it will refer to the the list of server provided in AutoStartList. You could assign minimum number of servers in the AutoStartList which could help as workaround to achieve your concern temporarily.

However, as Gaurav said, there is no variable available as of now in the VCS which could address your request. You can provide your idea at - http://enhancement.symantec.com/

Regards
Rajesh

Note: Please mark this note as solution if this helps you.

 

 

Regards

Rajesh Regunta

---------------------------------------------------------------------------------------------------------------------

PS: Please mark this note as solution, if this helps.