Hi Thomas,
Ahhh, that is a difficult task, I have to say...
I just finished a project with IS-U implementation with BW reporting, and reconciliation was a pain in the ¤¤¤ with the IS-U invoicing. The extractor is so complex, picking up data from all over and processing it on the way, so reconciliation was extremely dififcult.
I mostly used two different methods.
a) I created a view combining ERCH, DBERCHZ01 and DBERCHZ03 tables to get header, qty and pricing data together. That one I used a lot during the project phase, however, you can't simply reconciliate 1-to-1 data with infocubes using that view. Anyway, that was a good tool in all kinds of error scenarios, but you would need to sort the view in many ways, when you use it. I normally sorted using company codes, billdocs and st.groups. Remember to exclude the non-booking relevant lines. I also made a different view where I added ERDK (with several tables in between ERCH/ERDK), but that was so complex that it wasn't performing with large amount of data, and it wasn't always reliable in complex scenarios, where old migration data was combined with current invoicing.
b) The CO transaction KOB1 was actually a better way of reconciliating data between cube and IS-U. However, it depends on the customizing how the data is reported in your IS-U in KOB1 and how the transfer/update rules are set in your BW. Because the reconciliation was regarded as a highly important matter, I actually changed the transfer/update rules in our BW so that we were able to match the KOB1 results (by business unit and posting date) with the results in our cube. Our BW structures were originally based on business content, but they were anyway highly modified. That's the process customer is continuing to use today and are satisfied.
Hope this gives you some idea.
Regards,
Lauri
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.