Clustering Enterprise Vault in a 2+1 configuration without "any-to-any" support

Figure: 2+1 configuration without "any-to-any" support illustrates a configuration in which there is a single spare node in addition to the two nodes on which the Enterprise Vault servers are running.

Figure: 2+1 configuration without "any-to-any" support

2+1 configuration without "any-to-any" support

If either NODEA or NODEB fails, the virtual Enterprise Vault server running on that node can fail over to NODEC. This is not an "any-to-any" configuration so if a node fails the resources must be moved back after the node is recovered, in order to return to high availability.

To cluster Enterprise Vault in a 2+1 configuration without "any-to-any" support

  1. Prepare for clustering as follows:

    • Add three nodes to the cluster (NODEA, NODEB, NODEC).

    • Create two resource groups (EVSERVER1, EVSERVER2), and add the prerequisite resources to each group.

    • Configure the groups and resources so that the following nodes are the preferred owners, in the order shown:

      EVSERVER1

      NODEA, NODEC

      EVSERVER2

      NODEB, NODEC

    • Create DNS entries for the virtual servers EVSERVER1 and EVSERVER2.

  2. Install Enterprise Vault on NODEA, NODEB, and NODEC, but do not run the Enterprise Vault Configuration wizard.

  3. On NODEA, run the Enterprise Vault Configuration wizard and choose to configure a new Enterprise Vault server with cluster support. Select EVSERVER1 as the resource group in which to create the Enterprise Vault service resources. A Vault Site alias will be created automatically using the virtual server alias.

  4. On NODEB, run the Enterprise Vault Configuration wizard and choose to configure a new Enterprise Vault server with cluster support. Select EVSERVER2 as the resource group in which to create the Enterprise Vault service resources. A Vault Site alias will be created automatically using the virtual server alias.

  5. On NODEC, run the Enterprise Vault Configuration wizard, and choose to configure a failover node for the existing clustered server. Select either EVSERVER1 or EVSERVER2 as the resource group for which you want to add this node as a failover node.

  6. Test the cluster to confirm that if NODEA fails, the EVSERVER1 resources fail over successfully to NODEC. Then return the EVSERVER1 resources to NODEA and confirm that if NODEB fails, the EVSERVER2 resources fail over successfully to NODEC.