| || |
Input data from a data source
I am testing a client server ODBC application (SQL Server). I need to be able to have the contract number of a contract that can be in one of three states, active, pending, or cancelled. For some transactions, I need an active contract. For others, I need a pending contract, etc. I can get this information from the SQL Server database, but I obviously don't want to skew the performance results.
I could grab the data from the database beforehand and put contract numbers into three different files, but I'd have to do that before almost every test. One transaction, for example, will cancel an active contract. That contract number in the active contract data file is now invalid.
I thought about an excel spreadsheet as well. I could get the data once and have the transactions update it as they change their status. The spreadsheet could be on a network drive and I could make ODBC calls to update it. What would that do to the performance numbers? Are there any other good ideas?
I'm just going to create the data files for now. I'm dealing with over 1 million contracts in the system, so the chances of picking the same contract twice are almost astronomical. I still wanted to ask for future projects that will come up.
Re: Input data from a data source
Use 3 arrays instead of files? It'd be faster.