cancel
Showing results for 
Search instead for 
Did you mean: 

VVR with Geo-cluster Implementation.

Hello,

We are planning to implement a DR solution using VVR for data replication and Geo-cluster for failover.  The production side consists of 2 node cluster (2 ldoms) and the DR consist of 1 node cluster (1 ldom).  On the production side we have 1 ldom per customer with a single VCS service group for the customer, whereas some of of the ldoms have several customers in a single cluster with multiple service groups (1 service group per customer).  We need the ability to failover a single customer to the DR site.  In the case of single ldom or enviorment per customer we can do this easily. 

My query is in case of multiple customers/service groups on the single ldom and in a single VCS cluster, will this cause any limitation or is just  case of failing over the service group for the particular customer.

2 Solutions

Accepted Solutions
Accepted Solution!

You are correct ,this is just

You are correct ,this is just the case of failing over the service group for the particular customer - i.e you can fail each service group to the DR independently of the other service groups.

Mike

View solution in original post

Accepted Solution!

Hi, when you say multiple

Hi,

when you say multiple customers, depends on how you are  configuring multiple customers.

Option 1 - If you are configuring multiple service groups with in same cluster -each service group for each customer & you can configure each service group as global. In this way, you will keep each customer independent of other as resources are in different service group

Option 2 - If configuring a customer means configuring an application resource with different mount points & IP address, you can all club them in one single large service group where each resource will need to have carefully defined dependencies. Also the failover policy of group or which resource needs to be critical etc these things will need to be planned carefully.

G

View solution in original post

2 Replies
Accepted Solution!

You are correct ,this is just

You are correct ,this is just the case of failing over the service group for the particular customer - i.e you can fail each service group to the DR independently of the other service groups.

Mike

View solution in original post

Accepted Solution!

Hi, when you say multiple

Hi,

when you say multiple customers, depends on how you are  configuring multiple customers.

Option 1 - If you are configuring multiple service groups with in same cluster -each service group for each customer & you can configure each service group as global. In this way, you will keep each customer independent of other as resources are in different service group

Option 2 - If configuring a customer means configuring an application resource with different mount points & IP address, you can all club them in one single large service group where each resource will need to have carefully defined dependencies. Also the failover policy of group or which resource needs to be critical etc these things will need to be planned carefully.

G

View solution in original post