The debate between whatever

This is going to be heavy on linking to tech opinion articles.  I have tried to keep my eye on tech trends for many years.  About 15 years ago I would pic up something like PC magazine and scan through it.  I still remember the first time I saw a WTF crazy article, it was written by John Dvorak.  To sum it up, he claimed that smartphones were bound to fail due to the high price and risk of losing them in a cab.  This was around the time of the Palm TREO, so he even got smart phones wrong again with his iPhone article years later.  His viewpoint is interesting, but he gets a whole bunch of things wrong about technology and trends.  For the most part, it is entertaining to read his articles.  The articles that I have seen popping up lately though, faulty logic and a disturbingly low understanding of tech and integration are on full display. 

Aggregate 1:
Dvoraks infamous iPhone article

Starting a few years ago, I noticed things like "the debate between virtualization and containers".  I had to take a step back and wonder what poorly constructed AI wrote this miserable article.  It was equal to a garbage post on a comment section.  The authors always seemed to be bringing up specific examples of how and why containers work better on platform X, which makes me wonder about their kickbacks.  Container hosts are generally going to be virtual machines.  Calling containers a VM killer is hyperbolic and does nothing to educate or enlighten the reader.  It is also ignoring all development trends with the biggest companies in both businesses.  I went about my business, creating VMs and not caring about containers until the tech was a little more mature.  It took about four years before I really started getting serious about using them in a professional environment.

Aggregate 2:
Containers vs Virtual Machines!!!

So, now that I am using them in a professional environment, the next big thing that we can quickly dispose of is "will the cloud kill your data center?".  These articles are always fun to read.  In some instances, yes, you will migrate to a nice secure cloud and your company probably will not save money when they do it.  But currently, I am not concerned about clouds destroying all data centers.  They are not as good at certain tasks and protocols that are requirements for a significant amount of companies, for now.  I'm not kidding, go sit down with AWS and ask how they handle multicast in their offerings.  There are also tools to migrate workloads to the cloud when necessary, once again making the technology a complimentary one.  So, while this is an actual threat to some data centers, it is not nearly the end of data centers around the world. 

Aggregate 3:
Cloud killed the on-prem star

My most recent example of ridiculous articles was based on a really cool offering from Amazon.  I have written about the use of AWS Lambda in my echo controlled remote article.  It is amazing technology and an excellent example of cloud services.  The sales pitch I got from AWS was that this is "serverless" technology.  A fancy way to say that they have isolated your code in an environment that does nothing but run scripts when necessary.  It has been referred to as "black box" technology because Amazon won't spill the beans on exactly how they do it.  It's a clever method to tie your code to a container running whatever language your script was written in.  This is where the term "microservice" becomes important. 

Your cloud based application, a container, probably has an engine that works on a specific long running task.  Serving a web page, monitoring a device, maybe a database engine.  But, for a specific event, there should be a script that runs.  In traditional computing, that script would run locally and you would have to worry about patching the application to run it.  Maybe that script offloads data to another application or service, which traditionally would reside on the same box or container.  With current cloud architecture, the correct flow for this would be Container A --> Lambda --> Container B.  You have broken a series of applications apart and linked them back together with a Lambda.  You can now scale horizontally to improve your application, which means you only need to turn on additional instances of whichever container is getting slammed.

This is obviously a complimentary technology.  I apologize for the long buildup on this, but I really wanted to frame this appropriately before I share the article.  Because, to me, this article is the equivalent of "RAM vs CPU, which should you put in your computer?".  To make matters worse, this is from a company that is trying to sell DevOps-as-a-Service.  It is their job to know about these things and explain them correctly. 

Aggregate 4:
containers-vs-serverless

No matter how much I want to write things like "John Dvorak is a crazy guy that shouldn't be taken seriously" after reading some of his articles, I am yet to find an instance in where he attempted to sell a service (other than as a commentator) based on his point of view.  I get it, people are trying to make money off of open source technology and vendor offerings.  They need to appear educated about what they are selling.  It doesn't help Caylent that they have massive errors on their "devops-as-a-service" sales page, and then present a testimonial from someone that is happy about how they set up his WordPress.  Not kidding, as of 06/15/18, they have the blurbs under "Built to Grow" and "Automated Disaster Recovery" mixed up.  This is a problem, one that involves credibility. 

Think about that whenever anyone tells you that some new technology is competing when it could actually be complimentary.