When search requests and search replies cannot be delivered immediately to a node, VTAM® queues the data for delivery when possible. In situations where a node is not responding and searches are accumulating rapidly, VTAM storage can be excessively used. To prevent this overuse of VTAM storage, the MAXLOCAT start option can be specified.

The value of the MAXLOCAT operand is compared to the product of the number of queued searches and the amount of time between responses from the node. If the product is greater than the MAXLOCAT value, VTAM issues message IST1601I, which indicates that routing to this node is suspended and VTAM has stopped sending searches to the node until the congestion is reduced.

Even after VTAM suspends sending APPN search requests to an adjacent CP, VTAM continues to queue APPN search replies to the adjacent CP. If the number of queued requests plus the number of queued replies reaches 200% of MAXLOCAT and a minimum time interval has been exceeded, VTAM issues message IST2243I indicating that VTAM will initiate termination of the CP-CP session to the adjacent CP. VTAM ends the CP-CP session with sense code 80030004 and tries the CP-CP session again. This allows the storage related to the queued APPN searches to be cleaned up.

The minimum time interval is based on how long it has been since VTAM issued message IST1601I indicating routing to this node has been suspended. If IOPURGE is less than the default of 180 seconds, then the minimum time interval is 180 seconds. Otherwise, the minimum time interval is the current IOPURGE value.

When coding a value for MAXLOCAT, consider the amount of VTAM storage available and the response time of nodes. For example, set the value high if there is plenty of available storage and a slow node exists.