Entry 70 still missing

[Former Member]

Hello Experts,

I have browsed these forums on the matter at hand, but was unable to find a solution to my problem, hence my post.

We are retreiving data from our XI system. This is done via Web Service (Push) adapter (data is being pushed to BI from XI).

While launching a test packages into BI, we have come across a strange situation: the data is being saved to PSA (and is physically present there) almost instantly but the process, when checked in the monitor, is still "Yellow" and a message

Data Package 1 : arrived in BW ; Processing : Processing not yet finished, entry 70 still missing

can be found in the Transfer (IDocs and TRFC) part of the Details tab of the process monitor.

I've read on SDN that this is because the process still continues due to large amount of data and I can let it roll and it should eventually complete (turn "Green') or I could manually set the status to "Green" myself - however - in our case, we are talking about 130 records taking... 18 days 20 hours to finish (sic!) - at least that's what the header of the process monitor is stating, and I don't see why it would lie

While I could understand that large volumes take some time to finish loading properly to BI, I am disoriented as to why it would take almost 3 weeks to load 130 (one hunder and thirty) records.

Could anyone point me in the direction of the probable cause of this situation? I would be deeply thankful.

Best regards,

Aleksander

[Former Member]

This error usually happens when try to load a cube with the Index ON. You will need to drop the indexes in the F-table before you load or Usually this error comes up when ther multiple requests populating to an single data target.One requests restricts the entry of another into the data target.Wait for the present request to be updated and then re-run the next request.

To fix that error, try these steps:

- In the monitor (TA RSMO) select the "red" job in the left-hand frame.

- Go to the menu and select "Edit >>> Init. Update >>> Setting for further update".

- A dialog box will appear. Select the option "Packet should be processed in the background".

- Go to the menu again and select "Edit >>> Update/Reversal >>> Read everything in manually".

- A dialog box will appear with the message "Incorrect data packets are now updated again". Click on the green check-mark. This will submit a batch job in SM37 for each failed data packet.

- In SM37 the job name(s) will start with BI_BOOK.

- You can also monitor the status by selecting the "Details" tab and clicking the "Refresh Details" button.

OR

This could be problem of TRFC struck. TRFC's are struck due to some RFC connection problem between the source system and target system. To come up with this, go to the SM58 Screen, over there U may find some tRFC's having status " Transaction Recorded " (or) with some " Error " statement. Select that specific tRFC's 1 at a time and goto Edit in the Main Menu, there U will find the option Execute LUW( F6 ), just select that option, after that try and refresh the screen thru refresh button, U will not find that tRFC in the list. It will be succesfully posted to the Target system.

Hope this answers your issue...

-Vikram

[Former Member]

Hello all,

Thanks for your suggestions, however:

@Martin - no logs or short dumps are available either on BW or XI side (since - as I have said - the process does not terminate, merely takes unsually long to properly finish)

@Vikram - as I have said, the problem is realted to s PSA, no InfoCube. TRansaction SM58 (tRFC maintenance) contains no information and gives no help

I have managed to bypass this issue by creating a process chain and including a special step in it - "Close Request of Infopackage (RDA/Push)" - at the beginning. I have defined infopoackage from the discussed datasource in it and run it some time after the push has been done from the XI system - as I have said, physically the data is being present in the PSA in a matter of seconds / minutes, so it's only an issue with the technical processing of the package(s).

The problem with my solution is that I can never be 100% sure that the data is in BW, but by providing a long enough time window between pushing the data into BW and running the aforementioned chain, I can get sure enough to sue it.

[Former Member]

Hi ,

I am also facing the same issue.Could you please let me know the procedure for solving the issue if you would have already resolved it

Regards

Ram.

[Former Member]

Hello Ramakanth,

Please check my post above to see how I dealt with this issue (there's a slight typo at the end, it should read "I can get sure enough to use it").

Edited by: Aleksander Dyl on Feb 22, 2010 2:39 PM

[Former Member]

Hello Experts,

While loading data from a Flatfile to the DS through IP, I am getting the error “ Entry 70 still missing” and the load is getting stuck every time.

There are around 1 Million records that I need to load from the flatfile. I tried with lesses no. of records ( a few thousand ), but again the same problem.

I have tried the TA RSMO solution mentioned above but the option – “Update/Reversal >>> Read everything in manually” is disabled for me.

Any other suggestions to solve this problem? Please help.

[Former Member]

I believe you mean- whenever BI request data from ECC or XI it establish a connection first and then data transfer get started. In your case you were not getting the job completed confirmation for your source even after the job had completed in XI. We can check the issue with Status job.