Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

Disable AutoFailOver from stopping services

Created: 18 Jul 2012 • Updated: 23 Jul 2012 | 4 comments
This issue has been solved. See solution.

I have several services that I am monitoring that are set to autofailover to a seconds system. It has some up that occasionally I need to restart a service without HA failing over to another system. I can disable the failover from happening using the following command
hagrp -modify App_Cluster AutoFailOver 0

However what happens is that if the service is stopped HA continues to shutdown all the other services that are up. I was researching and I cam across disabling the Evacuate on HA, but even with it disabled, it still shuts down the other services.
hagrp -modify App_Cluster Evacuate 0

I want the other services to continue to run even if one went down for some reason. What is the best way to accomplish this?

Comments 4 CommentsJump to latest comment

mikebounds's picture

If you occasionally need to restart a service without HA failing over to another system, then you should freeze service group:

hagrp -freeze App_Cluster

VCS will still then monitor all the components in App_Cluster, but if it notices one is down, it will not take any action to "clean" that component or offline other components.

When you have restarted component, you can then unfreeze service group:

hagrp -unfreeze App_Cluster

If you permantently don't want VCS to take action, you should set VCS resource to non-critical, but this will only work if the resource is at the top level in the resource tree - i.e no other resources depend on it.

Mike

UK Symantec Consultant in VCS, GCO, SF, VVR, VxAT on Solaris, AIX, HP-ux, Linux & Windows

If this post has answered your question then please click on "Mark as solution" link below

mkruer's picture

While trying to freeze the cluster, I am getting the following error

VCS WARNING V-16-1-10150 Group App_Cluster cannot be frozen when it is transitioning in the cluster.

This is the hastatus, i don't see anything wrong

group resource system message
--------------- -------------------- -------------------- --------------------
app-49-207 RUNNING
app-49-209 RUNNING
App_Cluster app-49-207 ONLINE
App_Cluster app-49-209 OFFLINE
-------------------------------------------------------------------------
app1 app-49-207 ONLINE
app1 app-49-209 OFFLINE
app2 app-49-207 ONLINE
app2 app-49-209 OFFLINE
app3 app-49-207 ONLINE
-------------------------------------------------------------------------
app3 app-49-209 OFFLINE
app4 app-49-207 ONLINE
app4 app-49-209 OFFLINE
app5 app-49-207 ONLINE
app5 app-49-209 OFFLINE
-------------------------------------------------------------------------
Virtual_IP app-49-207 ONLINE
Virtual_IP app-49-209 OFFLINE
Network_Card app-49-207 ONLINE
Network_Card app-49-209 ONLINE

mikebounds's picture

You need to run "hastatus -sum" - this should show a resource than is onlining or offlining.  If it is onlining, then you can run "hagrp -flush grp_name -sys sys_name", but if it is offlining then you will need to stop the component manually (outside of VCS control)

Mike

UK Symantec Consultant in VCS, GCO, SF, VVR, VxAT on Solaris, AIX, HP-ux, Linux & Windows

If this post has answered your question then please click on "Mark as solution" link below

SOLUTION
mkruer's picture

Just out of curiously the step required in order to restart a service would be
1. freeze the group
2. stop the app manually via "service app stop"
3. hares -clear app -sys system_name
4. unfreeze the group
5. hares -online app -sys system_name

Any reason why i would not just use the offline app instead and then online it?
that seems to work just as well and does not require all the other steps