Typical Friday


I got handed a bunch of database functions a while back by my customer. He told me to get them to work and deploy them. It was a rush job. I fixed some bugs and got the things to compile. Then I did a few tests and fixed some functional problems. Then I pushed them out.
 
Ever since then, problems have been coming up with the functions and I get pulled in to resolve them. The goal was to hand these functions over to another team to maintain. But the other team is busy and says they know nothing about the business behind the functions. Well neither did I…
 
This past Friday, somebody discovered that one of the output tables from the function was empty. The team who was supposed to assume control of the functions was called in to investigate. They had no clue what the problem was. So I got brought in to make sense of the discrepancy.
 
At first I just came up with the direct cause of the situation. The records in one source for the function did not match the records they are supposed to be lined up with. Therefore, nothing was output. That was not a sufficient answer.
 
Of course this was some type of emergency. The valid output was due to some testers a while back. As a work around, I pumped some test data into the output table to give the testers something to look at. Then I dug in to find what was wrong.
 
The source table with unmatched records was quite small, thousands of records instead of millions. So I tracked down what was the problem. Apparently it was loaded with some test data instead of the real deal. There you go. Problem solved. Now there was a reason that the real data could not be loaded in, but that is a whole other story for another day.