How can I verify performance with program in sproxy?

[Former Member]

Hi,

I have an interface that when execute a file take like two hours to process the file... The only way that I have and I know to debug and check the program is using the transaction SPROXY using a local file for the test..

Exist a way for example to use transactions like ST05 or another one to check in what part of the code I have the performance issue??

Thanks a lot for the help!!!

Carlos

[Former Member]

Hi,

Thanks for the help!

I explain the scenario:

I have an interface that when receive some files the execution time is too long... this is an inbound interface...

To verify the performance issue I go to the transaction SE80 to verify the class method (EXECUTE_ASYNCHRONOUS), over there I see the ABAP code so I put a breakpoint in some part of the code... for example in a loop and then I call the transaction SPROXY to execute the process with the file... so I go to SPROXY, I select the software component version in Message Interface (inbound) and the I click execute, when I execute it the system show me a pop-up where I select the check button XML Editor, after that I upload the file that I want to use for the test and I execute it to begin the debugin process.

This is the only way that I know how to check the performance for a interface... I'm not sure if exist other way....

[udo_martens]

Hi Carlos,

unforturnatly you did not describe your scenario. I guess it is flat file -> ABAP proxy. If so you can look to the time stamps in Communication Channel Monitoring of Runtime Workbench (PI Java Stack), SXMB_MONI at PI (PI ABAP stack) and SXMB_MONI in ERP.

Usually in such scenarios the PI mapping is the performance bottleneck.

Regards,

Udo

[former_member182503]

Carlos,

are you talking about a FILE to PROXY interface, right?

If so, you can encapsulate the program logic of your INBOUND PROXY into a function module (like Z_PROCESS_FILE_IN).

Then, you can create a dummy program (like Z_RUN_TEST) that reads the file and convert it to the input parameters of your function module and executes this function module in the end, simulating the whole scenario of your interface.

After all, to measure the performance bottleneck, use the transaction SE30 using your dummy program as parameter.

Best regards,

JN