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

Configuring MSCS cluster resources for VVR

Created: 14 Feb 2012 • Updated: 15 Feb 2012 | 3 comments
This issue has been solved. See solution.

Hi Forum,

I've got a question regarding the configuration of cluster resources for VVR within a Windows Server 2008 R2 Enterprise 2-node-failover-cluster (clustered application is File Services).

The "Veritas Storage Foundation and High Availability Solutions 5.1 SP2 Solutions Guide" (page 823 et seq.) describes that you have to set up the RDS between an IP address of one node of the primary cluster and one node of the secondary cluster. Then you have to set up the cluster resources, including a client access point with a virtual IP. This resource has to fail over to the second node on the primary resp. on the secondary site of the VVR in case of a manual or disaster triggered application fail over.

I setup the resources according to the guide. Here's the setup:

Primary Site:

Primary Cluster Node 1: 172.128.128.1/24

Primary Cluster Node 2: 172.128.128.2/24

Cluster Resource CAS: 172.128.128.3/24

Secondary Site: (without cluster, just a single server)

Secondary VVR Server: 172.128.128.4/24

RDS:

Setup between 172.128.128.1 and 172.128.128.4

 

- The replication between Primary Cluster Node 1 and Secondary VVR Server works perfectly.

- The failover of the cluster resources works fine.

- Replication between Primary Cluster Node 2 and Secondary VVR Server doesn't work, they even can't connect to each other.

 

For me it is very clear that Primary Cluster Node 2 and Secondary VVR Server cannot replicate the RDS because it is configured between IP addresses of Primary Cluster Node 1 and Secondary VVR Server and not  between Cluster Resource CAS and Secondary VVR Server. I tried the latter but it seems to be impossible.

Does anybody know where's the rub?

 

Thanks in advance!

Regards,

Bob

Comments 3 CommentsJump to latest comment

Wally_Heim's picture

Hi Bob,

Basically, you need replication to use an IP address that will be in the service group that has the dynamic disk resource and the VVR resource.  This way the IP that replication is using will follow the service group and it will be available on any node that the group is running on.  The Secondary site since it is pointing to a virtual IP for the primary will follow the IP address when the primary is moved between nodes.

Here is more on your example.

On the priamry cluster you need to setup replication to use a virtual IP address that can failover between the two nodes at that site.  You should modify the RDS to replicate between 172.128.128.5 and 172.128.128.4. 

Then in the service group where you have the VVR resource add an IP resource that is for IP address 172.128.128.5 and make the VVR resource dependent on the IP resource just created.

Thanks,

Wally

unex's picture

Hi Wally,

problem was that the guide partly describes the setup for a one-to-one (primary-to-secondary site) and partly the setup for a cluster-to-cluster configuration - the latter is correct.

I don't know if there are any editors at Symantec, but the guide ("Veritas Storage Foundation and High Availability Solutions 5.1 SP2 Solutions Guide") to set up a MSCS with VVR (page 826 et seq.) should tell you first to create the cluster resources and then to create the RDS. Within the description of creating the RDS it should be corrected to use the virtual IP and not the IP of one of the cluster nodes.

Thanks and best regards,

Bob

SOLUTION
Wally_Heim's picture

Hi Bob,

Thank you for the feedback on the guide.  I've put in a request to our document writers to review the section of the guide that you mention.

Thanks,

Wally