cancel
Showing results for 
Search instead for 
Did you mean: 

Queue goes in Sysfail state. Pls advice

Former Member
0 Kudos

Hi,

When I send wrong xml to my scenario which has BPM

my queue goes in Sysfail state and shows message

"Permanent error in BPE inbound processing"

Pls advice urgent.

Regards

Accepted Solutions (0)

Answers (2)

Answers (2)

aashish_sinha
Active Contributor
0 Kudos

Hi,

Queues that correspond to the template XBQO$PE_WS* or XBPE_WS* are assigned the status SYSFAIL and error message SWF_XI_INBOUND 004: "Permanent error in BPE inbound processing".The inbound processing encountered a permanent error and the queue therefore stopped.

The error message "Permanent error in BPE inbound processing" may be caused by the following:

- At the start or when you process a process instance, the inbound processing encounters an exception that was not handled. This exception is triggered, for example, by a transformation step, a synchronous send step or a control step but it was not caught in an exception handler branch. If the exception that was not handled occurs in the first transaction of the started process instance, the system does not write the process instance to the database but it sets the queue that has this error message to the error status.

Refer to the help in Note 1082099.

- The inbound processing triggers this error if the correlation evaluation shows that a message should be delivered to a process instance that has the status "error". Since delivery to process instances that contain errors is not permitted, the queue is stopped.

Use the simulation of the message delivery in the detail view of the relevant queue LUW to determine the relevant process instance for the affected message and use transaction SXMB_MONI_BPE "Continue Process Following Error" (or transaction SWF_XI_SWPR) to continue. If you cannot continue the process instance, check whether you can logically delete the instance. In this case, use the method that is described in Note 807906 to logically delete a process instance.

Check whether your system contains other process instances that have errors in order to detect similar problems. Use transaction SXMB_MONI_BPE "Continue Process Following Error" or transaction SWF_XI_SWPR.

Use the report RSWF_CRL_MAINTAIN to check all of the correlation instances in question or the relevant process instances that you can reach using "Display Work Item".

- Messages that cannot be delivered because no correlation instance can be found may lead to the error that is mentioned here. Check whether the causes that are outlined in Note 1094028 affect your case.

-The inbound processing triggers this error if the system can locate a correlation instance for a message but cannot locate an active receive step for the message. However, an active receive step is required in accordance with the delivery mode "Without Buffering" (transaction SWF_INB_CONF).

Check the help in Note 1040614.

Check whether the relevant process instance may have terminated due to a system crash (for example, runtime error) and whether the process instance can continue so that it can reach a suitable receive step. Use the simulation of the message delivery in the detail view of the relevant queue LUW to determine the relevant process instance for the affected message. If required, use transaction SXMB_MONI_BPE "Continue Process After System Crash" (or transaction SWF_XI_SWPC) to continue the process. If you cannot continue the process instance, check whether you can logically delete the instance. In this case, use the method that is described in Note 807906 to logically delete a process instance.

Check whether your system contains other terminated process instances in order to detect similar problems. Use transaction SXMB_MONI_BPE "Continue Process After System Crash" or transaction SWF_XI_SWPC.

Use the report RSWF_CRL_MAINTAIN to check all of the correlation instances in question or the relevant process instances that you can reach using "Display Work Item" to see whether they provide the required receive step.

Hope this will solve your problem.

Regards

Aashish Sinha

PS : reward points if helpful

Former Member
0 Kudos

hi,

Do a complete cache refresh.. using sxi_cache transaction.

Regards,

Shabari