Configuring a Server as a Worker Node in a Cluster
For configuring a server as a worker node you need to set the CB_CLUSTER_SERVER_ID property as an environment variable. This property is normally used to configure the name of the node. You can use the same property to designate a server as a worker node by suffixing -background to it. For example,
CB_CLUSTER_SERVER_ID=<Node_1>-background
Once configured, the system administrator needs to enable the worker nodes using the backgroundServers application configuration.
Additionally, the
useOtherServers option within the
backgroundServers configuration allows the system administrator to set the time when other servers in the cluster can handle such jobs. For more details, see the
"backgroundServers" application configuration.
If a worker node is down, all pending jobs on that server are moved to other servers in the cluster based on the configuration in the backgroundServers application configuration.
Considerations when using a worker node for running jobs
• Any job that has already started on the worker node is completed on that server, even if the active user server is configured to execute the same job.
• Similarly, any job that has already started on the active user server is completed on that server, even if the worker node is configured for the job.
For information on roles of worker nodes in a setup, see
Load Balancer Configurations.