The Big Data Hype cycle is in full swing and I’m continuously hounded by vendors claiming to have an offering that works with Big Data. The problem is most of the vendors simply don’t get Big data. For example I am working with a security vendor that wants to offer a security solution based on the number of nodes. What they have a hard time wrapping their brains around is the fact that we will have thousands of nodes and their solution will only provide value for less than .001 percent of the data. Pricing by node is a non-starter. I have another hardware vendor that will go un-named that keeps coming out with hardware models that add more instrumentation on the mother board so I can see if ASIC 42 is dying, running hot or whatever. Hey we just added over a million dollar cost to your infrastructure that adds zero value to you…. In a Hadoop world this is worthless – I need a bunch of paper bicycles and any one of the servers can fail and we are going to keep chugging along. The worst of the “head in the sand” vendors are the storage vendors. They simply just can’t wrap their brains around storage at 20 cents a GB, JBOD, and the whole distributed compute/storage model. Their model is let’s just creating a big proprietary appliance that locks them in and makes scaling difficult and expensive. You can count on storage vendors losing massive market share as we adopt these types of technologies outside of Hadoop. With all that said – the worst of the worst are the open source vendors who somehow think they can continue tying support to hardware specs…you’d think they of all vendors would understand that won’t work in this model. Even the Hadoop distribution vendors are challenged and haven’t really adjusted to the new world. My normal vendors are driving me toward an Open Compute model and a totally non-subscription based open source model…thanks for helping me out of my rut.