7.7.21 Troubleshooting when executing the business process
While a business process is executing, the failures shown in the figure below may occur. This subsection describes how to handle these failures.
- Organization of this subsection
(1) Failure in an asynchronous receive activity
If an error occurs when a message is sent from the service requester to an asynchronous receive activity, the HCSC server is placed in a status in which it can no longer accept messages. The actions in such cases are as follows:
-
Eliminate the cause of error.
-
Resend the same message from service requester.
(2) Failure in a synchronous receive activity
To handle an error that occurs while a message is being sent from the service requester to a synchronous receive activity:
-
See the error message which is output in message log.
-
Eliminate the cause of the error (this is required for certain types of errors).
-
Resend the message from service requester.
(3) Timeout when sending a message to a synchronous receive activity
The following describes how to handle a timeout that occurs while a message is being sent from the service requester to a synchronous receive activity.
- Action
-
Since it is unknown whether or not the message sent from the service requester was executed, check if the error message output to the message log.
- When an error message is output
-
-
If an error message has been output, take appropriate action according to the error message.
-
- When no error message is output
-
-
Use the cscpiselect command to check the process instance status in the correlation set of the corresponding message.
-
If the process instance is still executing, take one of the following actions:
Re-execute the process instance using the cscpireexec command.
Delete the process instance execution log, using the cscpidelete command and, resend the same message from service requester.
If the process instance status cannot be acquired, reception of the corresponding message has failed. In this case, resend the same message from the service requester.
-
(4) Error during service invocation (Exception)
When you invoke service in the invoke service activity of business process, if the following errors occur, the execution of process instance will stop.
-
The service adapter invoked from invoke service activity is not running
-
The service adapter invoked from invoke service activity is not deployed
-
Data transformation process in the service adapter invoked from invoke service activity, has failed
-
Errors in the processing of synchronous service component invocation (errors such as invalid service component address, HTTP error due to stopping of service component, communication timeout, communication failure)
-
Errors in the processing of asynchronous service component invocation, when usage of business process compatibility is OFF (errors such as invalid queue address and failure in the queue send process)
-
Errors in the processing of DB access using DB adapter (errors such as DB access error, SQL error, invalid DB Connector settings)
-
The fault messages responded from Web service in the service invocation process of SOAP, does not satisfy conditions for handling as a fault in HCSC server
For details about the conditions for handling fault messages returned from service components as faults, see "4.3.5 Catching a fault with the SOAP fault operation definition file" in "Service Platform Overview".
Note that the status of the applicable invoke service activity, remains Error or Executing, when the usage of business process compatibility is OFF (bp-status-compatible in HCSC server setup definition) and; remains Executing when the usage of business process compatibility is ON.
Following are the steps to be taken in case an error occurs:
-
Eliminate the cause of error in the service adapter or service components, according to the error information.
-
Re-execute the process instance with the cscpireexec command and invoke the service component again. In this case, you will be invoking service component with the same message for the second time. Therefore, service component must be able to perform duplicate check and process the second call without any conflict.
(5) Fault from service components during service invocation
If a service component responds with a fault message when HCSC server executes synchronous execution, perform the following:
- Steps
-
- When fault connection is defined
-
Process is executed as per the defined fault connection.
- When fault connection is not defined
-
The process instance is placed in fault termination status and stops executing. Service component resends any messages that do not result in a fault respond.
- Important note
-
You cannot re-execute a process instance with the cscpireexec command, if SOAF Fault returned from service component during service invocation, is a fault. This is because the status of business process does not remain as interrupted, as business process is executed as per the defined fault connection and process instance execution stops.