Batch trace in ST12, schedule trace leads to different system.

[sujit_sharma]

Dear Experts,

I came across a situation where I've to schedule a trace of background job. In ST12 when I'm providing all required selection and as soon as I click on "Schedule Trace" it leads to session manager of a different system.

To be more specific- The system where I'm scheduling the trace is QA refresh of ECC landscape, when I click on "Schedule Trace" it opens the session manager ECC Productions system's one of the dialog instance.

Checked all the RFCs, Logon Load balancer, RFC load balancer . I could not find the system anywhere. Checked all profiles present in the system too.

Is there any other possibility which can lead to this behavior ?

Best regards,

SUJIT

[Reagan · Accepted solution]

Hello

How about SM61 ?

Regards

RB

[sujit_sharma · Accepted solution]

Hello Reagan,

Haven't created any group either for batch jobs in QA.

Thanks,

SUJIT

[Reagan · Accepted solution]

Hello Sujit

Check whether the system is using the correct profiles.

Import the profile from the OS level - Tx RZ10

Cross check the profiles under Tx RZ04 - instances/profiles and restart the SAP system.

Check if that helps.

Regards

RB

[sujit_sharma · Accepted solution]

Hi Reagan,

These all things seem to be fine.Have attached the screenshots from SMLG/RZ12/RZ04 below:

Have raised an OSS message for the same,but SCN help is always quicker than OSS specially for those with priority as "medium".

Hope with you all I'll be able to sort it earlier than SAP gets involved. Please let me know if any other idea we can have to sort this out.

Thanks,

SUJIT

[Former Member · Accepted solution]

Hi Sujit

From your QAS system . do you have any RFC destination pointing towards the PROD system , especially the 7th dialog instance ? 

Thanks

RD

[Former Member · Accepted solution]

Hi,

As i told to you in the earlier blog so should be related to one of the things in those transaction.

What are these profile...shown in RZ12 and SMlG.

What is 01 and 00 instance..

Can you please try to delete the config in SMLG snf RZ12 and try to reconfigure it for central instance only for QE1 system and then same it .

Then try to run the trace..

Thanks

Rishi Abrol

[sujit_sharma · Accepted solution]

Dear Rishi,

There is no RFC connection created or exists for PRD system.

Thanks,

SUJIT

[sujit_sharma · Accepted solution]

Hi Rishi,

Those profiles are related to QA,Central instance (00) and Dialog Instance (01). Have already tried deleting and creating again the logon goup/rfc groups however as you suggested I'll try once doing the same having assigned to Central Instance only. This is the only thing not done yet.

Thanks,

SUJIT

[Former Member · Accepted solution]

Check technical settings for hossapdia01_QE1_01 RFC destination in SM59.

[sujit_sharma · Accepted solution]

Hi Rishi,

Tried your suggestion, situation remains the same. Waiting for SAP to come back.

Thanks for your replies, Appreciate the help extended.

Regards,

SUJIT

[Former Member · Accepted solution]

Hi,

Can you please check the BD64 if you can see any reference of production.

Thanks

Rishi Abrol

[sujit_sharma · Accepted solution]

Dear All,

Many thanks for your efforts, it seems that problem is some kind of strange. SAP is working on it and even they are having hard time. Got a reply from them yesterday:

-------------------------------------------------------------------------------------------------------

03.02.2014 - 20:43:46 CET - Info for Customer by SAP
Hi Sujit, I find the piece of code where the call to hossaped07_PEC_00 is made, but no correction for it. I sent this message to our development level.

----------------------------------------------------------------------------------------------------------

I'll update the solution as soon as it is provided.

Best regards,

SUJIT

[sujit_sharma · Accepted solution]

UPDATE- Below is the information from SAP:

------------------------------------------------------------------------------------------

As a workaround, when you choose ST12->Schedule trace, check whether

you find any remaining old requests from system PEC in status

"Job found" or "Trace active". If yes then put the cursor on the line

and choose button "Stop trace". This should hopefully remove the line

before the trace demon is restarted.

For the next ST-A/PI release 01R* we will a check that eliminates

invalid remebered server namen in scheduled trace requests.

----------------------------------------------------------------------------------------------

Best regards,

SUJIT

[Former Member]

Can you describe in detail which steps (preferably with screenshots) you performed before receiving session manager screen?

[sujit_sharma]

Hi Roman,

I'm trying to schedule a trace in QA system for a background job by scheduling the trace, when I'm clicking on Schedule trace it opens the session manager of a Production system's one of the dialog instance, we have 10 dialog instances but it goes to one specific (7th dialog instance).

Please find the screenshot attached :

Nowhere could find the relevant entry for this system in QA.

Thanks,

SUJIT

[Former Member]

Sujit

Are you getting the same thing when you are trying for workprocess ? Or only for batch job ..

can you please check your SM61 background job server configuration ..

Thanks

Rishi

[sujit_sharma]

Rishi,

It happens only while scheduling the trace as shown in the image posted earlier. Rest all is fine.

Thanks,

SUJIT

[Former Member]

Can you provide screenshot from tcode rz12?

[Former Member]

Hi Sujit

One way to find out the cause would be to put a ST05 trace on your ID ( select all traces ) ...

and then perform the activity and check the resulting trace..

It will lead you to the place from where the RFC call is being made and the PROD  screen is opening up ...

Thanks

Rishi

[Former Member]

Hi,

Can you please pass on the screen of rz12 ,rz04 and smlg .

Did you run BDLS correctly.

Do you see any thing in BD64 and BD54 related to production.

Thanks

Rishi Abrol

[sujit_sharma]

Hi Rishi,

Thank you for replying, I had tried this and it shows the correct system information only there. It doesn't show the other system at all there.

The system was refreshed nearly 10 months back and all the entries related to PRD were corrected. Wondering how this started, developers had taken traces in batch till 2nd Jan of this month.

Any other idea ?

Regards,

SUJIT