University graduation (not mine) I have mixed feelings on learning that there is now a university-level subject in Performance Testing. CS 4803 EPR: Performance Engineering – Enterprise Computing Performance Engineering, presented by Brian Wilson at the Georgia Institute of Technology looks like a great subject, and it would be a lot of fun to do what Brian has done and run the same course at a local university. But, at the same time, I realise that doing this would be a clear case of fouling your own nest.

Flooding a profitable niche industry with recent university grads who are prepared to work for practically nothing and can claim practical experience with specialised tools will only push salaries in one direction. Obviously these people won’t be getting lead positions, but filling the bottom end of the market will, in turn, put pressure on the senior people who do get lead positions.

But this is really an inevitable trend. As the performance testing industry matures, people with performance testing skills will become more of a commodity. This is a good thing for software quality, but not such a good thing for people who are taking advantage of the current skills shortage.

That said, I am yet to meet a performance tester who was hesitant about sharing their knowledge. Maybe this just means that (collectively) we value good software more than high salaries.


Published On: November 8, 2005Tags: ,


  1. Scott Barber November 11, 2005 at 7:47 am

    You’re kidding, right?? PLEASE tell me that you are kidding!!

    1) I can’t tell you what I’d give to hire up a bunch of these kids! I could almost always use 1 (or 15) sharp, not overpriced scripters with a clue! If you’re really concerned about your profit margin, just hire them!

    2) You can’t seriously think that some recent grads with 13 weeks of “experience” will actually compete with professional performance testers can you?!? I honestly don’t see them even competing with the 80% of people who claim to be performance testers who are classic “fools with tools”.

    3) Even if every university with a Computer Science program offered this course, I doubt it would be manditory and I further doubt that more than… say… 1% of the students who take this course will actually end up specializing in performance testing. I certainly wouldn’t call that a flood!

    4) Cem Kaner and I have been trying to find the time to build a similar course for over a year. Trying to boil just the critical concepts down to a 13 week college course leaves a LOT of room for growth. I’m just about convinced that it would take most of an Associate’s Degree to really develop competitive performance testers right out of school.

    5) If I get driven out of my profession, or if my bill rates go down do to an influx of recent grads with a single performance testing course, then I never deserved what I was making in the first place and it serves me right for taking advantage of my clients.

    I, for one, am thrilled to see this. I’ll be volunteering to be a guest speaker, assist with course improvements and providing the instructor with “For Academic Use” versions of virtually all of the training and/or informational materials that I own the rights to. Maybe it really will generate a new generation of folks that can grow into high quality performance testers in 3-5 years as opposed to the 10 it has taken most of us who have earned a clue or two.

    Scott Barber
    Chief Technology Officer
    PerfTestPlus, Inc.
    Better Testing… Better Results

    “If you can see it in your mind…
    you will find it in your life.”

  2. Anon November 15, 2005 at 10:04 pm

    Sure, those things take years to master. But that’s mastery. Does it take years to learn how to play chess? No. Does it take years to become a master of chess? Yes.

  3. Jeff Jewell February 2, 2006 at 5:14 am

    I can understand Stuart’s concern, but my own concern would be more along the lines of the dangers of a little knowledge. I know that there are lot of companies who want to do performance testing of their systems but aren’t willing to pay for an experienced person.

    What I see happening is that people who have taken a class in college will get these jobs and do a poor job of testing the systems. That won’t help them or the company they’re working for.

    However, I welcome the existence of any formal training on software testing, and I’d love to take the course myself. Most of my knowledge comes from the streets and I’m sure I could learn things to make me better. Any training I could get to keep me from making the kinds of mistakes I’ve made in the past would be welcome.

    By the way, I just discovered your blog. Thanks for making the effort.

  4. Alex Podelko March 22, 2006 at 1:33 pm

    found an interesting link about the subject:

  5. Stuart Moncrieff March 22, 2006 at 10:05 pm

    Thanks Alex. From your link, it looks like there are at least 6 other university subjects with a focus on computer performance.

    …oops, one link is dead with a 404, one course is now offered by e-learning, and one course hasn’t been run since 2003.

    I will have to check out the course materials for the other ones though.


  6. Muhammad Abubakar November 19, 2006 at 6:57 am

    let me get details information through my email

  7. Brian Wilson December 7, 2011 at 2:41 am

    Hi all. It’s been over 6 years since I taught this course (one semester only) at Georgia Tech. It was a lot of fun, but I had to travel and could not continue to commit to being in town every Tue and Thu, so I had to make it a one shot deal.

    I’m more or less in the same situation now. I travel 3-4 days a week. However, if I thought there was enough interest in some ideas I’m mulling around in my head, I may rearrange my schedule sometime in the future.

    Would any of you be interested in the following subjects if the course(s) were conducted remotely in the late afternoon/evenings and were very interactive? Perhaps you have clients who would be interested and/or you may be interested in doing so presentation yourself.

    Please contact me at bwilson at techsouth dot com if interested. I don’t have any immediate plans right now, but would like to develop some consultants, thus the interest in doing training in some of these areas. I am setting up a PC11/ALM rig that will be RDP-accessable, as well as an SAP environment also RDP-accessable. If there was enough interest in any of these areas, I’d create the course. I have lots of content from which to start, but productising learning is a time-consuming endeavor.

    – Performance Test Planning: Entire course would just be about PTP planning best practice with lots of real world examples, artifacts, etc. This would include PTP project/resource managing.
    – Test Monitoring: All about what to monitor, how to monitor and why to monitor. I would probably use .Net and SAP for interactive examples
    – LoadRunner Scripting for SAP: GUI and EP (Enterprise Portal) scripting including a team project using my environment. Most of the time will be spent of EP, of course. Those who excel will certainly get project offers from me. ; ) Must have very solid scripting skills to enter this coure; mastery of advanced techniques and concepts will be assumed.
    – Value Added Results Analysis: Root cause analysis of complex results not only using LR Analysis, but using other tools and techniques.

  8. Brian Wilson December 7, 2011 at 2:44 am

    I forgot one more course:
    – Using PC11: Designed for performance testers and PC administrators. This course is already fully developed and has approx 21 hours of content including labs. No scripting, just using PC11 and administration of PC11. It is great for those moving from PC 9.x to 11 as well as LR to PC.

Comments are closed.