The human mind is a brilliant classification engine.  We interact with the world around us and quickly classify everything we meet across a variety of spectrums…friend or foe, good or bad, fight or flight, dangerous or safe, fun or boring, interesting or dull, ugly or beautiful, young or old, happy or sad….you get the picture.
Each end of each spectrum has a known set of parameters on why it would be one versus the other, but the real beauty of the human mind is that it can run these classification jobs in parallel across a variety of spectrums for any particular situation and comprehend that while we are using binary classifiers, the true nature of anything may be somewhere in the middle of a lot of these spectrums.  We are pretty awesome like that.
My mind over the last few years has been consumed (at least professionally) with the classifications that I, my customers, and our industry place on the deployment options for modern analytics frameworks like Hadoop and Splunk.  So when I was I asked to develop a talk aimed at helping executives and IT leaders understand architecture concepts for these Big Data technologies, it made sense to me to juxtapose a number of the most popular spectrums I see our industry consider when looking for the right way to deploy Hadoop.
What I have found is that the idea of using spectrums juxtaposing big ideas is much more useful that having finite, hard rules or responses to architectural questions.  The main reason?  Because the right answer for architecture and deployment sort of depends…on a number of requirements, technological phenomenon, and situational realities.  I won’t categorically say that any one concept is more right than another, rather I sought in this session to provide reasons why a situation or organization might chose one versus the other based on the value or opportunities in doing so.
And because I work for Dell EMC, who has the largest portfolio of product offerings for Data Analytics in the industry, I have the benefit of talking with folks about every option in this architectural catalog of concepts, not just one or two where we have products to fit…we can generally meet customers and provide value wherever ever they sit across a variety of spectrums when it comes to modern architecture for Big Data.  Please enjoy this talk where we investigate Modern Architecture Concepts for Big Data Deployments across a unique variety of classification spectrums.
Screen Shot 2017-06-22 at 6.38.02 PM
And you can grab the slides here.
Your Bearded Friend,
PS – “juxtapose” is almost as fun to say as “chimichanga”