A common question I have been asked by customers over the years after they have upgraded their systems is, “Where did my capacity go?” This can be a difficult conversation to have. As I was researching other items, I came across a Wikipedia entry on Jevons’s Paradox. It states, “In economics, the Jevons paradox (sometimes Jevons effect) is the proposition that technological progress that increases the efficiency with which a resource is used, tends to increase (rather than decrease) the rate of consumption of that resource.” It got me thinking that this could be applied to this situation.
When we conduct Capacity Planning on a Teradata system, we look at workloads based on defined user groups, and their historical consumption rates. We tend to see organic growth rates from about 2% - 5% on a monthly basis. This includes both an increase in the underlying data volumes, as well as new user adoption of these capabilities. We also forecast expected new workload being introduced and it’s take up rate, and growth. These are the primary contributors to understanding capacity on Teradata and are captured, measured and reported on through the Capacity Planning and Performance Tool. The missing, and much harder factor to measure is pent up demand.
This also leads to another Wikipedia entry: Parkinson's Law “Work expands so as to fill the time available for its completion.” Or further generalised as, “The demand upon a resource tends to expand to match the supply of the resource.”
Let’s say you are an analyst in an organisation. One way or the other, you will be putting in your 8 (or 10) hour day looking for the next great insight in your business. Before the upgrade to the system, you are able to run 3-4 analytic queries a day, review the information, revise and refine, and work your way towards the answer you were seeking. Then you come in one Monday morning, and see your queries finishing much quicker than before. Yes, the system has been upgraded with more capacity, and that extra capacity results in more throughput. So now, instead of running 3-4 analytic jobs, you can increase your work to 6-8 jobs per day. The more questions you ask, and the more answers you get, lead you on and on down the path.
Applying Jevon’s Paradox enhanced by Parkinson’s Law – upgrading your system will only serve to increase demand as the efficiency improves and more work will be done in the same amount of time.
So, where did your capacity go? It went to the users!
Next up… Part 2: Addressing Demand and Capacity