Jon has made performance and scalability testing the focus of his career since 1988. He has completed hundreds of engagements encompassing thousands of tests against applications from every major industry and then some. To him, technology is a playground to which the rules of the game are constantly changing... Jon works out of the Arizona office (360+ sunny days a year; sorry east coasters) and lives in Goodyear with his wife and college son (Go ASU).
Thank God You're Here
Thought I'd start off with a little thought dump. I just watched a few episodes of "Thank God You're Here", a comedy show where actors are dressed up and thrown into a scene with absolutely no knowledge of what they are getting into and have to improvise in front of an audience and home viewers until a judge has had enough and stops the scene. Some of the actors do quite well, stealing the scene and making it work (mostly through comedy). Others get thrown for a loop and struggle through it (gets boring). This is not too far from my world. There are largely two situations I find myself in. First is when a company has performance testing as part of their pre-deployment in which case my job is to help them “make it work.” Second is when a company encounters a performance problem and calls upon RTTS to help them resolve the problem as quickly and efficiently as possible. That’s the “Thank God You’re Here!” This is the time when we (RTTS) are thrown into the scene, and find out if we have the knowledge and experience to make it work. I’m going to write about many things here, including experiences of mine, perplexing and philosophical questions, methodologies, interesting occurrences, issues and resolutions. I hope some of what I do here will elicit responses and participation or all to share.
Posted by Jonathan Harris on Thursday, May 31, 2007 3:12 AM EDT
Leave a comment
| 
RE: Thank God You're Here
"Thank God You?re Here" is exactly how we feel every time we engage RTTS. Performance testing is part of our pre-deployment strategy. Demand for testing far exceeds staff resources. "Amen" to functions like "list of values" and ?random number generator? that are needed to automate the requirements of many test cases. Record and Replay will get you to about 70%. It is the other 30% custom coding that really makes the tool perform the desired task. Furthermore, this 70/30 relationship varies with the protocol. Additional challenges arise when the tool does not work quite the way you expected. (Can I get an "Amen" from the LoadRunner folks out there?) Thanks again RTTS. You guys are the best.
Posted by P. Stokes (Pepsi Bottling Group) on Sunday, April 20, 2008 8:54 AM EDT