How network definition is affected by multiple systems

Each system definition must have all terminals that are connected to it defined during system definition using either the TERMINAL and NAME macros or defined dynamically using the type-2 commands CREATE LTERM and CREATE MSNAME. However, if terminals in other systems are capable of sending messages that are destined for the defined system, those terminals are logically part of the network for that system.

You can define your MSC physical and logical links either dynamically with CREATE MSPLINK and CREATE MSLINK commands, or statically through the system definition process. You can define your MSC remote LTERMs either dynamically with the CREATE LTERM command, or statically with the stage-1 system definition NAME macro, or during IMS initialization with Extended Terminal Facility (ETO).

When a transaction is processed in a remote system, the input LTERM name in the local system is carried over as part of the message. If the processing program uses the alternate PCB to direct a message to another terminal besides the input terminal, those destinations need to be declared as remote, unless directed routing is used. Define the LTERM names for all inputs with NAME macros or with CREATE LTERM commands. Position the NAME macros in a group after the MSNAME macro. You now have a set of LTERMs that collectively can occur in several system definition decks. For example, TERMA can be present in the input system, in the intermediate system, and in the processing system.

When planning for the network, keep in mind that message queues for an input system or an intermediate system must allow for the remote transactions being queued. When allocating space for the message queues take into account the message lengths and their expected loads. In a similar way, you must allow for the presence of these messages in I/O buffers, even when they are not going to be processed in that system.