Running Task Engine in a Clustered Environment
Depending on the runtime on which Task Engine is installed certain requirements and considerations apply for using Task Engine in a clustered environment.
For Task Engine on My webMethods Server:
If you are working in a clustered
Integration Server environment, all
Process Engines in the cluster typically share a single instance of
webMethods Broker (deprecated) or
Universal Messaging, a
Process Engine database component, or both. In this case you can connect from
Task Engine on
My webMethods Server to any
Integration Server/
Process Engine in the cluster. Communications will be established with all
Process Engines in the cluster.
If you are working in a clustered
My webMethods Server environment, you must apply the
Task Engine server role to each server node in the cluster, as described in
Administering My webMethods Server. This role must be applied to all nodes of the cluster where
Task Engine operations are performed; for example, tasks being worked on (either from an inbox or Task List Management) or queued from the
Process Engine. If there are nodes in the cluster where these operations would never happen, this role can be omitted from these nodes.
For Task Engine on Integration Server:
When
Task Engine is running on
Integration Server, you use a separate
My webMethods Server/
Business Console node to administer tasks and provide task user interfaces.
Task Engine registers as a
My webMethods Server cluster node and requires
Universal Messaging for JMS communication and event synchronization.
When creating a clustered environment, you must add
Task Engine instances, running on either
Integration Server or
My webMethods Server. You cannot create a cluster of
Task Engine instances, running on different servers.