Best way to look for missing sequencial numbers in infocube?

Question

I'm looking for some ideas on the best way to do the following:
The characteristic transaction number is in a cube by day, store, and register.  I need to ensure there isn't a missing transaction number across days...
Example:
5/16        Store:  5     Register: 1       Txn:  12345
5/16        Store:  5     Register: 1       Txn:  12346
5/17        Store:  5     Register: 1       Txn:  12348
(So in the example above txn 12347 is missing.)
We are talking about 300,000 transactions per day
First Idea:
I was thinking about using an infopackage to extract the 2 days from the cube(600,000 records) and I could use ABAP in the start routine to figure out if there is a missing txn.  Would that work?  I vaguely remember something about a start routine only having access to a package, not the whole dataset.
Second Idea:
Is there a way to have a query get the last txn # from previous day and the first txn # from the current day?  Then I could use a workbook with VBA to figure out the missing txns.
5/16        Store:  5     Register: 1       Txn:  12346
5/17        Store:  5     Register: 1       Txn:  12348
5/16        Store:  5     Register: 2       Txn:  67890
5/17        Store:  5     Register: 2       Txn:  67891
5/16        Store:  6     Register: 1       Txn:  45678
5/17        Store:  6     Register: 1       Txn:  45679
(This would be 1500 records since we have 150 stores, 5 registers each, for 2 days.  150 * 5 * 2)
Any other thoughts, ideas, or comments?
Thanks,
Chris

Answer 1

Not sure if the following method is the most intelligent,  but worth a try if you have no better way...
1) Create a query with required variables (like Register, Calendar Day etc) placed in the free characteristics and the Txn Number placed in the rows. If the Txn number is a characteristic, from the properties, choose SORT ASCENDING. Else you can use listcube t-code and sort the output in the ascending order, then save  the list as a local file (excel sheet) onto the desktop.
2) Create a separate Excel file with the transaction numbers in ascending order (assuming that you know the txn number range for those days in the period of interest). This is very easy to do in Excel especially since they will be in order (type 1 txn and drag it downwards which will increase the number serially).
3) Now that you have two files (first one is the as-is records and  the second one ideal records) compare these two files using a software like <a href="http://www.scootersoftware.com/beycomp_062808.exe">Beyond Compare</a>. I have used this software before and found it to be useful.
Not sure if this helps, but good luck!