Performance Zen

Author: spierzchala

  • New GrabPERF Features

    Added two new features to the GrabPERF system this morning.

    1. Display the Arithmetic and Geometric Mean by aggregated hour. This uses the datapoints in the ‘data’ table, which reaches back 10 days.
    2. Display the Success Rate by aggregated hour. Uses the same data as above.

    These are attached to each individual test screen.

    Have fun!

  • Chris Anderson talks to Jeff Bezos

    In this interview, Chris Anderson (of the Long Tail fame) speaks to Jeff Bezos of Amazon about that very topic.

  • Thoughts on Web Performance Excellence

    In writing the last post, I was thinking about what factors go into making the Web performance of a site “excellent”. What defines in the minds of the sites users/customers/visitors/critics/competitors that the performance of a Web site is excellent?

    These are usually judged by the standard factors:

    • Responsiveness
    • Availability
    • Traffic
    • Reliability
    • Security
    • Clarity

    But within the company itself, how is the performance of their Web site judged to be excellent?

    Right now, most people use the external metrics mentioned above to determine excellence. However, it must be remembered that there are two other critical factors that need to be considered when managing a large IT infrastructure.

    • Ease of Management. This is a metric that is often overlooked when determining if a Web site is excellent from an internal perspective. Often it is simply assumed that running a large IT infrastructure is incredibly complex; in most cases this is true. However, is it too complex too manage efficiently and effectively? How much time is spent finding the cause of problems as compared to resolving them?
    • Cost of Operation. This is always a big one with me. I look at sites that are trying to squeeze as much performance and availability out of their sites as they can. At some point, the business has to step back and ask, “How much does another half-second of speed cost us?”. When this context is placed around the “need for speed”, it may open a few eyes.

    When this two critical internal factors are combined with the raw external data that can be collected, collated and analyzed, some other ideas come to the forefront as KPIs in Web Performance Excellence:

    • Cost Per Second. The cost of a Web site is usually only calculated based on the negative metric of how much it costs when the site is down. Well, how much does it cost when the site is up? Can that number be reduced?
    • Revenue By Speed. Which customers spend the most on your site: LAN, home-broadband, or dial-up?
    • Person-hours per day. How many person-hours per day does it take to manage your Web site?
    • True Cost of Performance Issues. When there is a performance issue, the cost is usually associated with lost revenue. Reverse that and ask how much did it cost in time and materials to resolve the issue.

    The creation of new Web performance excellence metrics is crucial if companies truly want to succeed in the e-business arena. Business management has to demand that IT management become more accountable to the entire business, using metrics that clearly display the true cost of doing business on the Web.

  • Heavy-Tail Distribution in Web Performance Data

    I found a great example of a Heavy-Tailed Frequency distribution in my performance data today.

    heavy-tail-data-apr132009

    This clearly shows how data in the wild can be distributed in a non-normal fashion. In this case, there is a very heavy weight on the end of the tail, not simply a few straggling outliers.
    It has become very unusual to find sites that exihibit this degree of heavy-tailed behaviour  over the last year. When I started in this industry, this was more the norm than the exception.

  • Ideas on Evolving Excellence

    A good read on evolving excellence through simplicity.

    The concepts that apply in the manufacturing process also apply in Web performance. Simplicity makes Web performance excellence easier to achieve.

    What does your Web performance process look like?

  • Jeremy Wright Adds More On His Firing

    Jeremy has added more on his firing (here).

    I am going to burn my bridges with more than a few potential Canadian employers by saying that this does not surprise me at all from the Canadian management mentality. In some ways, it is still stuck in the Victorian era: paternalistic and vindictive. My interviews with Canadian companies have always left me going, “I don’t want to work for them!”.

    Why? Because the HR Teams at Canadian companies are designed to remove critical thinkers and free spirits. I have yet to find a Canadian company of any size where innovative thought and inventive concepts were allowed to flourish.

    I would name some examples, but that would get me into even more hot water.

    Corporate Canada — and Corporate America, for that matter — has to accept that people talk about companies: to friends; colleagues; and the world. Blogging just makes that more global.

    As Jeremy states, companies need to understand that blogger and companies have to agree where the line is, and soon.

    Hmmm…more than a few hits on this article from HSC, Jeremy’s former employer.

  • Jeremy Wright Fired for Blogging

    Jeremy Wright was fired for blogging. This is likely the trend of 2005.

    I like Jeremy and as a fellow Canadian, this doesn’t surprise me. But I will hold my tongue on this topic.

    More here and here.

  • HTTP Standards Exist — USE THEM!

    I am extremely steamed at an article that just read on Caillon’s Blog which basically encourages people to disable HTTP Pipelining.

    This is the wrong approach.

    If a server announces that it is HTTP/1.1 compliant, then  it should be able to handle a browser that is using all of the HTTP/1.1 features. If someone is using a server version which cannot handle all of the features of HTTP/1.1, then they should be forced to fall back to HTTP/1.0.

    By continuing to bow to the lowest common denominator, Web performance will not improve and Server developers will not be forced to accept that they must fix their code.

    The only reason that server developers have gotten away with lousy pipelining support is because MSIE still does not support it. If MSIE begins to implement pipelining, then watch the mad scramble to resolve this issue.

    Proactive Web performance excellence. Do it.

  • GrabPERF Upgrade: cURL

    The GrabPERF system has been upgraded to cURL 7.12.3.

    Let me know if you see any weird behaviour.

  • Stupidity at two levels

    Ok, I did two very stupid things yesterday.

    1. Left the power supply for my laptop at home
    2. Dropped a Unique Index on my database log table and ended up with a huge number of dupes in the table. Spent two hours teaching myself how to rebuild the table without the dupes.

    Ok, so I know a lot more about indices and how they work now.