August 12, 2015

The Tire Pressure Rule

Having witnessed many internal IBM EMM implementations and sizing exercises, I’ve seen organizations either go full overkill or be Scrooges while spec’ing out their marketing platform infrastructure. Many sizing outcomes were usually based on the production system execution load statistics at a given point in time. Sizing in this way ultimately resulted in poor system performance and or overall high system ownership costs.  This can be easily avoided by following, what I call, the tire pressure rule.

 

20150812-tire-pressure-guage-1459225-1280x960-1024x768

Think of your EMM infrastructure as a tire. The way it performs is based on tire pressure or the load it will be subjected to. Too little load / pressure, and you’re wasting gas and money. On the flip side, too much pressure and you risk a blowout when a load spike occurs. What the tire rule states is this: When filling up a tire and you don’t know the recommended tire pressure for the vehicle, fill the tire to the designated max tire pressure MINUS 5-6 psi.

Sizing marketing infrastructure is relatively the same. A buffer range of power balance needs to be determined to perform proper sizing. In order to do that, marketing organizations need to look at the current production system performance over a period of time. One business week is recommended. Record performance statistics like:

  • Storage IO
  • CPU / Memory utilization
  • Database loads

Like pressure, each statistic has high and low points. By extrapolating the performance statistics over a period of time, and NOT on a given point in time, you’re creating that tire pressure buffer to base infrastructure specifications on.

Campaign execution, like car loads, changes day to day. By using the tire pressure rule, your system will have the flexibility to easily handle solo cruising down the coast or the occasional rhinoceros hitchhiker to your nearest zoo.

Related Posts