Manual Restart Status in Inbound RFC call monitoring in AIF

[former_member582009]

Hi All,

I am working on a scenario where PI will call and RFC from ECC and update the data in ECC. On ECC side; I have created an RFC Function module In that function mmodule I am trying to do the processing through AIF by calling a method

This is the code I have written in the RFC FM and in the AIF configuration I have assigned an action FM which will do the processing. The problem is; Action FM is not getting triggered and in AIF/ERR transaction; messages are stuck in manual restart status.

Please  help me on this. Also Kindly let me know what are the customization steps required for this. Maybe I have missed something in customization.

Regards,

Pankaj Sharma.

[Former Member · Accepted solution]

Hi Pankaj,

two quick ideas:

1. Could it be that all AIF jobs failed due to a dump? This cannot be provided back to AIF and the message stay in status "in process". (check job (SM35) and ST22)

2. Did you call the method with a queue namespace and name? Is this set up in teh aif runtime configuration? Then you need to start the message processing with program /AIF/PERS_RUN_EXECUTE.

Good luck

Christoph

[former_member582009 · Accepted solution]

Hi Christoph,

Thanks for your valuable reply!

The reason was the first one mentioned by you (job failed) as I did not call the method with the queue namespace. I just passed the structure to it.

I checked in ST22 and saw the dumps and in  SM37 ; all the jobs in scheduled status. Then I manually processed the jobs by correcting the errors. Then it processed successfully.

Moreover I want to add that whenever we call the RFC from our system ; then system user will do the processing. So our processing does not stop in Action FMs (in case we have put a debugger there). So if any dump or anything occurs in between; its highly unlikely we can see the dump during processing. We have to check it in ST22. and if dump is not there then we should check SM37 for the scheduled jobs and user name should be the system user. If we find the scheduled jobs; we have to release the jobs.

SM37 screenshot : Highlighted the system user and the status (scheduled).

Thanks!

Pankaj Sharma.