cancel
Showing results for 
Search instead for 
Did you mean: 

Service group states

shiv124
Level 4

Hi ,

Env:LINUX VCS

when doing an activity i bought 2 non critcal resources offline .them my service group state was partial .Is there any diffrence in a state partial|online and partial ?

If we try to bring a critical resource offline the service group would failover to other node .But when doing a activity if we dont want the service group to be failed over we need to make the critcal resource to non-critical and start the activity ?

 

Thanks and regards,

siva

 

 

 

 

1 ACCEPTED SOLUTION

Accepted Solutions

Marianne
Level 6
Partner    VIP    Accredited Certified

If you use VCS to bring critical resource offline, it will not failover.
It will only failover if you use methods outside of VCS to take resource offline (like OS kill command or db shutdown command).

I don't have access to any cluster right now, but partial|online and partial is probably they same, depending on which command output you are looking at. 
Maybe show us what you see?

View solution in original post

3 REPLIES 3

Marianne
Level 6
Partner    VIP    Accredited Certified

If you use VCS to bring critical resource offline, it will not failover.
It will only failover if you use methods outside of VCS to take resource offline (like OS kill command or db shutdown command).

I don't have access to any cluster right now, but partial|online and partial is probably they same, depending on which command output you are looking at. 
Maybe show us what you see?

mikebounds
Level 6
Partner Accredited

I believe Marianne is correct - partial|online and partial is the same, depending on which command output you are looking at.

If you take a resource offline using VCS, then VCS will not failover the service group.  If you take the resource offline outside of VCS, then VCS will failover the service group if the resource is critical - if you don't want this to happen, then you should freeze the service group and then VCS does NOT take action (still monitors the state of resources). 

So when doing a activity if we dont want the service group to be failed, you should freeze the service group, rather than making resources non-critical as even if a resource is non-critical, VCS will call the clean entry point (tries to stop application forceably to make sure it is fully down and clean) if VCS detect the resource goes offline, which may interfere with the "activity" you are doing.

Mike

shiv124
Level 4
HI Mike and Marianne , I was using hagrp -state . hagrp -state|grep -i partial abc State abcd |PARTIAL| Thanks for the info I was in asumption that if a critical resource is offline/faulty (out of vcs and with VCS ) the service group will failover .