OpenTP1 Version 7 System Definition

[Contents][Index][Back][Next]

Appendix A.5 Application program shutdown

When a UAP terminates abnormally, the OpenTP1 automatically restarts the process of the UAP. However, the OpenTP1 shuts down a UAP if it has ended abnormally a number of times greater than a specified value so that system processing can avoid entering a loop of repeating a cycle of abnormal termination and restart. The shutdown of a UAP which has ended abnormally causes the client to return an error to the service requester intentionally, so that unnecessary reruns can be avoided.

UAPs that may be shut down in this way are SPPs (limited to SPPs of queue-receiving servers). SUPs, UAPs that handle offline work, and SPPs of server that receive request from socket will not be shut down.

The following explanation applies to SPPs, not to MHPs.

Organization of this subsection
(1) Relationship between shutdown and monitoring time
(2) Shutdown in each service group or each service
(3) User service definitions related to application program shutdown

(1) Relationship between shutdown and monitoring time

When to shut down a UAP which has ended abnormally is determined by the monitoring time and the abnormal termination count.

If the TP1/Server Base is in use, the OpenTP1 shuts down a UAP when the UAP has ended abnormally three times during a period of time equal to or shorter than the monitoring time specified in the system definition. If the TP1/LiNK is in use, the OpenTP1 shuts down a UAP when it terminates abnormally for the first time. Time monitoring for shutdown is not used.

If a UAP has terminated abnormally twice or less during the monitoring time, the OpenTP1 restarts the process of the OpenTP1.

Time monitoring begins when the first abnormal termination occurs. Even if the number of abnormal terminations is two or less, abnormal termination counting restarts when the second abnormal termination occurs.

The following figure shows the relationship between the monitoring time and abnormal termination count.

Figure A-6 Relationship between monitoring time and abnormal termination count

[Figure]

(2) Shutdown in each service group or each service

UAPs may be shut down at a time for one service group or one service. If shutdown for each service is specified and even when a UAP terminates abnormally three times during the monitoring time, only UAPs for the pertinent service are shut down. This means that other services belonging to the same service group can continue.

Shutdown for each service can be specified only for queue-receiving SPPs. Shutdown for each service group can be specified only when the basic facility of the OpenTP1 is the TP1/Server Base. For TP1/LiNK SPPs, only shutdown for each service can be specified.

(3) User service definitions related to application program shutdown

The following user service definitions relate to UAP shutdown. These user service definitions can be specified only when the basic facility of the OpenTP1 is the TP1/Server Base. When a UAP terminates abnormally, SPPs under the TP1/LiNK are immediately shut down if they have been working for the service involved in the shutdown.

Combinations of values assigned to the operands related to shutdown determine whether a UAP which has ended abnormally will be restarted or shut down. For details about the relationship between operand value combinations and the execution of shutdown, see the descriptions of the operands.