uCosminexus Service Platform, Overview
If you define any activity after the reply activity, you can execute that activity using Reliable Messaging or asynchronous EJB.
The function to be used for executing the activity after reply differs depending on the database and the existence of Reliable Messaging. The following table describes this concept as follows:
Table 3-12 Function to be used in execution of activity after reply
Reliable Messaging |
Database |
Yes |
No |
Yes |
- Reliable Messaging for persistent business process
- Asynchronous EJB# for non-persistent business process
|
- |
No |
Asynchronous EJB# |
Asynchronous EJB# |
- Legend:
- -: Not applicable.
- Note#
- The following functions cannot be used when you use asynchronous EJB instead of Reliable Messaging for executing the activity after reply:
- Recovery process that uses the retry function of Reliable Messaging cannot be used.
For details on the recovery process, see "3.4.7 Transactions of the reply activity".
- The following receptions and adapter that presume Reliable Messaging cannot be used.
Standard reception (MDB (WS-R))
Standard reception (MDB (DB queue))
Standard adapter (MDB (WS-R))
Standard adapter (MDB (DB queue))
- Organization of this section
- (1) Settings when you use asynchronous EJB for asynchronous execution
- (2) How to re-execute if an exception occurs in the execution of asynchronous EJB
(1) Settings when you use asynchronous EJB for asynchronous execution
When you use asynchronous EJB, set the following properties in the HCSC server runtime definition file.
- bp-reply-after-max-thread-pool-size
Specifies the maximum number of threads that can be generated in a pool. The default value is 32.
- bp-reply-after-min-thread-pool-size
Specifies the number of unused threads that can be stored within the pool. The default value is 16.
- bp-reply-after-thread-pool-keep-alive
Specifies the duration (unit: seconds) for which threads are stored till the end of threads, when the number of unused threads exceeds the number specified in min-thread-pool-size. The default value is 60.
For the settings of the HCSC server runtime definition file, see "HCSC server runtime definition file"in "Service Platform Reference Guide".
The following is an overview of the execution of an activity after reply.
Figure 3-54 Overview of execution of an activity after reply
(2) How to re-execute if an exception occurs in the execution of asynchronous EJB
The following is the status of processes when an exception occurs in the activity defined after the reply activity, or when an exception occurs at the time of executing asynchronous EJB. The processes are re-executed in a batch:
- Process instance: Executing
- Reply activity: (Completedwait)
- Activity defined after the reply activity: Unexecution
Depending on the re-execution of process instance, you can execute the activity defined after the reply activity.
All Rights Reserved. Copyright (C) 2015, Hitachi, Ltd.