Create backup clusters
Use this task to configure a backup cluster that handles EJB requests if the primary cluster fails.
Before beginning, create two clusters that are able to provide backup for each other. The objects and resources available in the primary cluster must also be available in the backup cluster. We must use the same cluster name, install the same applications, use the same application names, and define the same resources in the backup cluster as in the primary cluster.
The primary cluster and the backup cluster must reside in separate cells because a cluster must have a unique name within a cell.
Perform this task to create a backup cluster for your EJB clusters. When all the servers in the primary cluster fail, work is not halted because the backup cluster can continue serving requests for EJB work.
To configure a backup cluster, specify a name and a port. The port is called a domain bootstrap address and consists of a bootstrap host and port. The bootstrap host is the host containing the deployment manager in which the backup cluster is configured. The bootstrap port is equal to the bootstrap port for the same deployment manager.
The primary cluster and the backup cluster must reside in separate cells. The bootstrap host and port for the backup cluster determine which cell contains the backup cluster.
- Determine the bootstrap host and port of the backup cluster.
- Connect the console for the deployment manager containing the backup cluster.
- Click System Administration > dmgr > Ports > BOOTSTRAP_ADDRESS. The host and port for the BOOTSTRAP_ADDRESS instance is the host and port that the backup cluster uses. Remember these values for when you configure the primary cluster.
- Connect the console to the deployment manager containing the primary cluster. Click Servers > Clusters > WebSphere application server clusters > cluster_name > Backup cluster.
- Ensure that the name of the backup cluster is the same as the primary cluster.
- Click Domain bootstrap address. Specify the backup cluster deployment manager bootstrap host and port in the Host and Port fields. Click OK. The bootstrap host and port combined define a bootstrap address for the deployment manager. On the Domain Bootstrap Address page, use the Configuration tab to statically define the backup cluster; the static value is consumed each time the deployment manager starts. We can use the Runtime tab to define the backup cluster during run time only; when the deployment manager stops, the run-time backup cluster information is discarded.
- Click OK.
- Configure a core group bridge between each of the cluster core groups. Use an access point group to join the two core groups. In the deployment manager for the primary cell, configure an access point group that has a peer access point that refers to the core group access point in the backup cell. In the deployment manager for the backup cell, create an access point group that has the same name as the access point group created in the primary cell. Add a peer access point that refers to the core group access point in the primary cell. See the topic on configuring the core group bridge service for more information.
If we are configuring a V5.x cluster to back up a cluster that is on the current release, do not configure the core group bridge service. Core groups are not supported in V5.x. Therefore, the V5.x cluster does not belong to a core group. The backup cluster still functions using only the domain bootstrap address.
- Save the configuration.
ResultsThe backup cluster completes EJB requests when the primary cluster fails.
What to do next
If we experience problems when configuring your backup cluster, read the topic on what to do if the workload is not getting distributed.
- Backup clusters
Backup clusters mirror the primary server clusters. Mirrored cluster support is for EJB requests only.
- Backup cluster settings
Use this page to configure a backup server cluster. The backup server cluster is used if the primary server cluster fails.
Balancing workloads Configure the core group bridge service
Workload is not getting distributed