The purpose of the Fiscal06 report is to allow the user to review account totals based on sort/group and object rollup. The parameters in the report allow the user flexibility to create many different comparisons across models for specific account components. Each time our report developer, Susan, has to make a change to a report, she looks for ways to optimize the performance, always trying to make the reports more efficient. Part of the process of testing report changes is to measure the performance of the report.
I use a simple method to compare the performance of the new report with the old report. Using a test system with the old version, I run the report with default parameters, trying to get as many pages of data as possible. I click Go on the report while starting the stopwatch. When the report is completely presented for viewing, I stop the timing. This way the report averages should reflect the user experience with the report. I then install the new report files to the test system and repeat the process. After testing, I put the ‘old’ report files back in place on the test system – just in case anyone calls with a question we want the correct version of software in place.
To illustrate the incremental changes for Fiscal06, the average run time for version 10.02 was 1 minute, 19 seconds, returning 31 pages. After changes along the way, Fiscal06 in version 10.05 ran in 24 seconds with 41 pages of data. That’s much better, but version 11.02 completes in an average of 13 seconds, about 50% faster than version 10.05.