Under normal termination, no new service requests are accepted, and OpenTP1 terminates after completion of processing of all service requests currently in the scheduling queue.

If MCF is being used, OpenTP1 terminates after completion of processing of all messages currently in the input and output queues. Therefore, under normal termination, MCF monitors the input queue and then the output queue until no more messages remain. You specify the monitoring time in the timer definition (mcfttim) of the MCF communication configuration definition.

When the queues are being monitored, log messages (KFCA16532-I through KFCA16537-I) are output at specific intervals. Output of these log messages can be suppressed by mcfmsmsg (definition for suppressing the output of log messages) in the MCF manager definition.

Note that if new messages occur successively (or intermittently) in the input and output queues during normal termination, events might occur, such as normal termination does not end, the OpenTP1 system shuts down, or ERREVTA is activated.

Under forced normal termination, OpenTP1 terminates the same way as for normal termination in (1) above even if a server has terminated abnormally.

This termination mode allows the system to terminate normally even if a UAP server has terminated abnormally.

Since the system starts in the normal startup mode after forced normal termination, the last online status is not inherited.

Under planned termination A, no new service requests are accepted, and OpenTP1 terminates after completion of processing of all service requests currently in the scheduling queue.

If MCF is being used, all unprocessed messages in the input queue are processed. Therefore, under planned termination A, MCF monitors the input queue until no more messages remain. You specify the monitoring time in the timer definition (mcfttim) of the MCF communication configuration definition.

When the queue is being monitored, log messages (KFCA16532-I, KFCA16533-I, KFCA16536-I, and KFCA16537-I) are output at specific intervals. Output of these log messages can be suppressed by mcfmsmsg (definition for suppressing the output of log messages) in the MCF manager definition.

Note that if new messages occur successively (or intermittently) in the input queue during planned termination A, events might occur, such as planned termination A does not end or the OpenTP1 system shuts down.

Any unsent messages in the output queue are retained.

If processing of all messages in the output queue has not been completed at the time of planned online termination, OpenTP1 terminates under planned termination A.

Under planned termination B, OpenTP1 terminates after completion of the service that is being processed currently.

The scheduler processes only the service currently being processed without accepting any new service requests. Any other service requests are discarded.

If MCF is being used, all messages in the memory input and output queues are discarded. Any unprocessed messages in the disk input queue and any unsent messages in the output queue are retained.

Under forced termination, OpenTP1 terminates immediately without waiting for completion of the service being processed currently.

All service requests in the scheduling queue are discarded.

If MCF is being used, all messages in the memory input and output queues are discarded. Any unprocessed messages in the disk input queue and any unsent messages in the output queue are retained.

If an error occurs, OpenTP1 shuts down completely, which constitutes abnormal termination.

All service requests in the scheduling queue are discarded.

If MCF is being used, all messages in the memory input and output queues are discarded. Any unprocessed messages in the disk input queue and any unsent messages in the output queue are retained.

The following table explains the handling of service requests in the scheduling queue and messages in the input and output queues under the various termination modes.

Table 2-2 Handling of service requests in scheduling queue and messages in input and output queues

Termination mode	Service requests in scheduling queue	Messages in input queue		Messages in output queue
		Disk	Memory	Disk	Memory
Normal termination	P	P	P	P	P
Forced normal termination	P	P	P	P	P
Planned termination A	P	P	P	R	R
Planned termination B	D	R	D	R	D
Forced termination	D	R	D	R	D

Legend:

P: Processed entirely.

R: Retained.

D: Discarded entirely.