cancel
Showing results for 
Search instead for 
Did you mean: 

VVR Action Plan for DR failover activity

allaboutunix
Level 6

Hello Team,

 

I need a help  to create in high level plan for DR Drill failover activity for VCS GEO cluster which have VVR replication running on Production site.

We will have to failover the Service groups from prod site to DR site.

So, do we have to move RVG resources in DR site  as well OR Should we failover only required application SG?

COuld you please suggest  full action plan on this, requesting to provide information on urgent basis, please.

Thank you,

5 REPLIES 5

allaboutunix
Level 6

Please help, to address the above query.

Thank you.

frankgfan
Moderator
Moderator
   VIP   

Hello,

In order to "... to failover the Service groups from prod site to DR site.", you need to "manually fault a critical resource which cannot be recovered by the local HA during your site faiover test, like a network failure.

When the "failed critical resource" is not able to go online on the local cluster, VCS will offline all the resources in the GCO (grobal cluster service group) including disk groups and RVGs.  Once all the resources in the GCO is offline, VCS will online the GCO on the DR site per your configuration.

If this is a production cluster, make sure to organize a maintenance window which is long enough for you to conduct the site failover test.

 

For most large production system site failovers (drill or a real failover) , most of the times, the cluster/system admin's manual operation intervenes are required.

 

By the way, if the cluster runs mission critical applications, make sure to do a full data backup and also collect a fresh DataCollector output.

 

 

I have a query Frank in reference to your answer,

1. Do RVG will be online on Production and DR site simultaneously 

2. Suppose, we have to manually switch over the application SG in DR event to test the failover time then,

Do we need to failover the RVG service group  to DR node also?

We have RVG service group as Parent  and application SG as child group.

 

3. Could you please suggest a DR plan for this VCS/VVR GCO setup?

 

 

 

Anybody, could you please suggest @veritas Team

frankgfan
Moderator
Moderator
   VIP   

I replied your query yesterday.  Anyway let me do it again

Please see my answers/comments to your questions below beginning with "<<<"

1. Do RVG will be online on Production and DR site simultaneously 

<<< Yes and No.  if you only failover/switch the service group (parant) with the applications, the RVGS and the DGS will remain on line on both sites

2. Suppose, we have to manually switch over the application SG in DR event to test the failover time then,

<<< then the DGs/RVGs remain online on both sites if you have storage resiurces configured in a child service group

Do we need to failover the RVG service group  to DR node also?

We have RVG service group as Parent  and application SG as child group.

<<< Please consult with VCS/VVR user guides for service group dependencies.  there are many good sample configurations in the guides which you can use as "templates"

3. Could you please suggest a DR plan for this VCS/VVR GCO setup?

<<< there are many ways to conduct a DR test, one of them is to manually "fault a network resource (like IP/IPMultiNICB/MultNICB}.  when you do the site failover test, make sure that local failover will not work so VCS will failover to the remote (DR) site.