We have a 2-node VCS cluster at primary site running Oracle database on Oracle Sun SPARC servers connected to a NetApp filer. We have a third server for the DR which is connected to another NetApp FAS model storage. I have to configure VCS with NetAppSnapMirrorTypes and enable replication between filers. My questions are:
1. Should I be configuring Veritas GCO and using the WAC?
2. Should I be installing VCS on the third node and give it a unique name and just mention only one node name?
In continuation to my previous queries, I understand that I need to configure single-node cluster on the third server and use the single-node cluster as a stand-alone cluster without enabling LLT and GAB. After this, I need to enable replication between a 2-node cluster at primary and a 1-node cluster at DR. Please correct me if I am wrong.
No, you're not wrong.
What you're describing is a fairly common configuration and is fully supported. You should definitely configure GCO on both clusters to facilitate automated application failover beween the two sites.
You have two options :
1. Automatic failover to DR
> For automatic failover to DR, you'd need to configure VCS with GCO, which would require you to install VCS on the third node (DR host) as well. The third node installation will be a single-node VCS install (unique cluster name and single node name). However, when configuring it under GCO, the other cluster (the primary cluster configuration) information will be needed.
2. Manual failover to DR
> For Manual failover, need to not to configure VCS on the DR host. Just install/configure VCS on both the nodes on the primary site. For DR failover, you can online the disks manually on the DR node, after offlining the service-group on the primary.
In either case, what you mention is absolutely correct. You'll need to replicate the data from primary to DR.
If I understand the objective correctly ... you want to have replication between two filers which should be controlled by the VCS (something automated) ..
If that is the case, I would choose option 1 i.e VCS with GCO (yes with WAC) resource ... This will have global service groups configured which will enable global failover from one site to another site ..
secondly, yes, you also would need to install VCS on third (DR) node however the configuration will have additional steps of configuring GCO ...
Most imporant, you will need to have valid SFHA licenses (GCO license) to use GCO..
Information around GCO config can be found here: