Configuring the VVR agents for parallel applications
Prev Example setting up VVR in a VCS environment Next

Configuring the VVR agents for parallel applications

Use the RVGShared, RVGSharedPri, and the RVGLogowner agents to manage and monitor RVGs used by parallel applications in a shared environment.

Note:

Determine the node that is performing the most writes by running the vxstat command on each node for a suitable period of time; after you set up replication, specify this node as the logowner.

The prerequisites for configuring the agents are as follows:

Sample configuration files are located in the /etc/VRTSvcs/conf/sample_rac/ directory and include CVR in the filename. These sample files are installed as part of the VRTSdbac package, and can be used as a guide when creating your configuration. You can configure agents from the command line or from the VCS Java and Web consoles.

See the Symantec Cluster Server Administrator's Guide for more information.

To modify the VCS configuration on the Primary cluster

  1. Define two new service groups: A logowner group that includes the RVGLogowner resource, and an RVG group that includes the RVGShared resource replication objects.
  2. In the logowner group, define IP and NIC resources, used by the RLINKs for the RVG, and the RVGLogowner resource, for which the RVG and its associated disk group are defined as attributes.
  3. In the RVG service group, set up the RVGShared agent to monitor the RVG resource. Because it is shared, the RVG must be configured to depend on the CVMVolDg resource, which defines the shared disk group and its activation mode.

    Define the RVGShared and CVMVolDg resources within a parallel service group so that the service group may be online at the same time on all cluster nodes.

  4. Add the RVGSharedPri resource to the existing application service group and define the service group to be a global group.

    See the Symantec Cluster Server Administrator's Guide for instructions on how to create global groups.

  5. Move the CVMVolDg resource from the existing application service group to the newly created RVGShared service group.
  6. Set the following service group dependencies:

    • The RVG logowner service group has an "online local firm" dependency on the service group containing the RVG.

    • The RVG service group has an "online local firm" dependency on the CVM service group.

    • The application service group has an "online local firm" dependency on the RVG service group.

To modify the VCS configuration on the Secondary cluster

  1. Log on to a node in the secondary cluster as root.
  2. Ensure that all changes to the existing configuration have been saved and that further changes are prevented while you modify main.cf:

    If the VCS cluster is currently writeable, run the following command:

    # haconf -dump -makero

    If the VCS cluster is already read only, run the following command:

    # haconf -dump
  3. Ensure VCS is not running while you edit main.cf by using the hastop command to stop the VCS engine on all systems and leave the resources available:
    # hastop -all -force
  4. Make a backup copy of the main.cf file:
    # cd /etc/VRTSvcs/conf/config
# cp main.cf main.orig
  5. Use vi or another text editor to edit the main.cf file, making the following changes:

    • Edit the CVM group on the secondary cluster. Use the CVM group on the primary as your guide.

    • Add the logowner group and the RVG service groups.

    • Add an application service group. Use the application service group on the primary cluster as a pattern for the service group on the secondary cluster.

    • Since the service group is a global group, assign it the same name as the group on the primary cluster.

    • Define the ClusterList and ClusterFailOverPolicy cluster attributes.

    • Include the RVGSharedPri resource.

  6. Save and close the main.cf file.
  7. Verify the syntax of the file /etc/VRTSvcs/conf/config/main.cf:
    # hacf -verify /etc/VRTSvcs/conf/config
  8. Start VCS on all systems in both clusters.

    The application group should be online on both systems of the primary cluster.

    The application service group should not be online on the secondary cluster, but the CVM, RVG logowner, and RVG groups should be online.

Prev Up Next
Configuring the agents when VCS is stopped Home Configuring the agents for a bunker replication configuration