OTMA support for IMS-to-IMS communications

You can send OTMA messages from a local IMS system to a remote IMS system by using IMS-to-IMS TCP/IP communications.

To send a message to a remote IMS system, you define the remote IMS system as an OTMA destination in either an OTMA destination descriptor or an OTMA User Data Formatting exit routine (DFSYDRU0). IMS application programs running in IMS dependent regions can then insert messages to an ALTPCB with the OTMA destination name specified. OTMA queues the messages to a tpipe for IMS Connect, which delivers the message to the remote IMS system by way of the TCP/IP network.

You can use the OTMA client descriptor to set an ACK timeout interval that determines how long OTMA waits for an ACK or NAK response to a transaction message sent to a remote system.

OTMA messages destined for remote IMS systems flow one way only. To return responses that are generated by the remote IMS system to the local IMS system, the return path must be defined separately in the remote IMS system and the response must be returned as a separate transaction.

Local and remote instances of IMS Connect manage the TCP/IP connection between the local IMS system and the remote destination IMS system. The TCP/IP connection between the two IMS systems is defined in IMS Connect by using the RMTIMSCON configuration statement in addition to the other required IMS Connect configuration statements.

Use persistent connections for OTMA IMS-to-IMS TCP/IP connections to minimize the risk of an excessive number of tpipes accumulating on the remote IMS system. IMS cleans up idle tpipes unless output is queued to them.

You specify connection persistence to the local IMS Connect instance by specifying PERSISTENT=Y on the RMTIMSCON statement that defines the connection to the remote IMS Connect instance.