Shared channel recovery

Home

Shared channel recovery

The following table shows the types of shared-channel failure and how each type is handled.

Type of failure: What happens:
Channel initiator communications subsystem failure The channels dependent on the communications subsystem enter channel retry, and are restarted on an appropriate queue-sharing group channel initiator by a load-balanced start command.
Channel initiator failure The channel initiator fails, but the associated queue manager remains active. The queue manager monitors the failure and initiates recovery processing.
Queue manager failure The queue manager fails (failing the associated channel initiator). Other queue managers in the queue-sharing group monitor the event and initiate peer recovery.
Shared status failure Channel state information is stored in DB2, so a loss of connectivity to DB2 becomes a failure when a channel state change occurs. Running channels can carry on running without access to these resources. On a failed access to DB2, the channel enters retry.

Shared channel recovery processing on behalf of a failed system requires connectivity to DB2 to be available on the system managing the recovery to retrieve the shared channel status.

Parent topic:
Benefits

ic15460_

Home

Type of failure:	What happens:
Channel initiator communications subsystem failure	The channels dependent on the communications subsystem enter channel retry, and are restarted on an appropriate queue-sharing group channel initiator by a load-balanced start command.
Channel initiator failure	The channel initiator fails, but the associated queue manager remains active. The queue manager monitors the failure and initiates recovery processing.
Queue manager failure	The queue manager fails (failing the associated channel initiator). Other queue managers in the queue-sharing group monitor the event and initiate peer recovery.
Shared status failure	Channel state information is stored in DB2, so a loss of connectivity to DB2 becomes a failure when a channel state change occurs. Running channels can carry on running without access to these resources. On a failed access to DB2, the channel enters retry.