To reassign activities from a resource adaptor entity to a different node, use the following rhino-console command or related MBean operation, noting the requirements.

Note
Why reassign replicating activities?

A resource adaptor entity in the STOPPING state cannot return to the INACTIVE state until all the activities that it owns have ended. You can let a deactivating resource adaptor entity return to the INACTIVE state quicker by reassigning its replicating activities to other eligible nodes.
Warning When using the pool clustering mode, it is not possible to reassign activities from one pool node to another using this operation.

Console command: reassignactivities

Command

 
reassignactivities <entity-name> -from node1,node2,... -to node3,node4,...
  Description
    Reassign replicated activities of a resource adaptor entity from the specified
    nodes to other nodes

Examples

To reassign activities owned by the resource adaptor entity named sipra from node 101 to node 102 and 103:

$ ./rhino-console reassignactivities sipra -from 101 -to 102,103
Replicated activities for sipra reassigned to node(s) [102,103]
Tip
Reassigning to all available nodes

You can also specify an empty (zero-length) argument for the -to option. This reassigns replicated activities, distributing them equally among all other nodes that can adopt them (nodes on which the resource adaptor entity is in the ACTIVE state and the SLEE is in the RUNNING state).

To reassign activities owned by the resource adaptor entity named sipra from node 101 to all other eligible nodes:

$ ./rhino-console reassignactivities sipra -from 101 -to ""
Replicated activities for sipra reassigned to node(s) [102,103]

MBean operation: reassignActivities

MBean

Resource Management

Rhino extension

 
public void reassignActivities(String entityName, int[] fromNodeIDs, int[] toNodeIDs)
      throws NullPointerException, InvalidArgumentException,
            UnrecognizedResourceAdaptorEntityException,
            InvalidStateException, ManagementException;

This operation reassigns replicated activities owned by the named resource adaptor entity, on the nodes specified, using Rhino’s standard failover algorithm, to the nodes specified by the toNodeIDs argument. (If toNodeIDs is a zero-length array, the operation reassigns activities to any remaining eligible nodes.)

Requirements for reassigning activities

You can only reassign replicated activities from a resource adaptor entity to other nodes if the all the following conditions are satisfied:

  • The Savanna clustering mode is being used.

  • The node is a current member of the primary component.

  • The node is an event-router node (not a quorum node).

  • The operational state of the SLEE on the node is RUNNING or STOPPING.

  • The operational state of the resource adaptor entity on the node is ACTIVE or STOPPING.

Further, a node can only take ownership of replicated activities if it satisfies all the following conditions:

  • The Savanna clustering mode is being used.

  • The node is a current member of the primary component.

  • The node is an event-router node (not a quorum node).

  • The operational state of the SLEE on the node is RUNNING.

  • The operational state of the resource adaptor entity on the node is ACTIVE.

Also, non-replicated activities cannot be reassigned to other nodes, and a resource adaptor entity must end any non-replicated activities it created itself.

Tip You can choose to forcefully remove activities if a resource adaptor entity fails to end them in a timely manner.
Previous page Next page
Rhino Version 3.2