Using an integration node with an existing Windows Cluster (Windows Server)

You can use IBM® App Connect Enterprise with the existing high availability manager for Windows Server (Failover Cluster Manager).

About this task

For more information about the versions of Windows Server that are supported by IBM App Connect Enterprise, see the IBM App Connect Enterprise system requirements web page.

This topic summarizes how to complete the following tasks:
  1. Complete the prerequisite setup.
  2. Configure a local group.
  3. Create an integration node.
  4. Add integration node instances to the additional nodes.
  5. Add the integration node service to the cluster configuration
  6. Start and stop an integration node.
  7. Delete an integration node.

Procedure

  1. To complete the prerequisite setup, complete the following steps:
    1. Validate your Failover Cluster Manager configuration.
    2. Complete all of the steps documented in 'Supporting the Microsoft Cluster Service' of the IBM MQ documentation to create a cluster configuration that contains the queue manager on which you want the integration node to run.
    3. Note the haregtyp.exe command as a prerequisite for creating IBM MQ resources in the cluster.
  2. To configure a local group, ensure that the domain user under which you want the integration node to run exists in the local mqbrkrs group on all nodes on which you want the integration node to run.
  3. To create an integration node, use the following mqsicreatebroker command on the node on which your cluster is currently running.
     mqsicreatebroker INODE -q MQ1 -e E:\ACE\Workspace
    where:
    • INODE is the name of the integration node.
    • MQ1 is the name of the queue manager.
    • E:\ACE\Workspace is the directory of a shared disk (nonquorum) in your cluster configuration.
  4. To add this integration node instance to the other nodes in your cluster, switch your cluster to each node in turn. When a node is active, use the following mqsiaddbrokerinstance command.
     mqsiaddbrokerinstance INODE -e E:\ACE\Workspace
    where:
    • INODE is the name of the integration node.
    • E:\ACE\Workspace is the directory of a shared disk (nonquorum) in your cluster configuration.
  5. To add an integration node generic service resource to the cluster which contains the integration node queue manager, complete the following steps:
    1. Select the IBM App Connect Enterprise component integration_node service when asked. All other settings can be left unchanged.
    2. Add a dependency on the IBM MQ resource, to ensure that the queue manager is started before the integration node.
  6. To start and stop the integration node resource, use Failover Cluster Manager.
  7. To delete an integration node resource from the Failover Cluster Manager, take the following steps:
    1. Delete the integration node generic service from the cluster configuration.
    2. Use the mqsiremovebrokerinstance command on all nodes but one, moving the cluster between nodes before running each command.
    3. On the final node, use the mqsideletebroker command to completely remove the integration node configuration.

Results

For further information about configuring IBM MQ with a high availability manager for Windows Server, see the IBM MQ product documentation online.

See the following topics:
  • Introducing MSCS clusters
  • Supporting the Microsoft Cluster Service (MSCS)
  • Setting up IBM MQ for MSCS clustering