I tried writing templates, macros, and VBA scripting. After that it was a process of find and replace to get the test names right, making sure the number formats and highlights were correct in every row on every page, and general data cleaning.
No matter how I wrote my formulas they would almost always have to be adjusted, then tables and graphs would have to be adjusted depending on the number of tests, nodes, and various response times that were recorded during the project. Data had to be copied over, cleaned, sorted, and filtered.
I will not get into the tedious process of getting data from our in-house tools to Excel but it was a multi-hour process for me and I'm very proficient with Excel. My tool of choice for writing and delivering these reports was Excel, and why not? I come from a strong background in Excel and it was my belief at the time that I could make it do anything. We do have some in-house custom tools that can aggregate most of the data, but not all of it, and the reports generated in those tools are not able to be customized to the extent they need to be for the particular type of testing I do.
All of that data has to be summarized into a final report of test results that gets posted to an internal Jira project page to be consumed by the software development team. For one of the projects I was involved in we had more than 20 systems that data had to be collected from.ĭepending on how long a project runs for, there can be upwards of 50 tests that have to be reviewed, analyzed, and reported on.
No test is limited to a single machine, most of the time it's a combination of Windows and Unix devices (physical and virtual), end-user systems and servers. We test new and updated software before it goes out to our clients and we test everything: Windows and Unix CPU and memory, Java Virtual Machine performance, response times on both backend services and the User Interface, Database performance, the impact on our current software, network performance, and pretty much anything else you can think of. The short version of what I do is software performance testing. The data sets were also small enough for it to handle.Īlmost a year ago today I started working in Cerner's Abilities Lab as a System Engineer in software performance testing. Due to security reasons, I didn't have access to VBA, Python, or any other type of automation. Excel, for the most part, worked well there and it was the only option. I've used it for everything from keeping track of my personal budget, expenses, and workouts to multi-page personnel strength analysis reports in the Army. It’s great for organizing data, performing calculations and analysis, and even works as a simple database.
I’ve been using Excel since 2001 and until just a few months ago it was my tool of choice for data organization, analysis, and visualization. It probably drove him a little nuts but I can’t help it I love data analysis and I love Excel.
Last week, my team lead sent out a list of employees who hadn’t updated their project checklists, and I immediately opened it up, started making tables and getting summary statistics by employee, then pinged him some summary data, broken down by employee. I’m the type of person who manipulates data in spreadsheets just because I can.