Wow has it been a whirlwind over the last ten years in the virtualization space. Where once Xen and then KVM sat on the pedestal, the baton has been passed to the projects revolved around containers. Names like Docker, Kubernetes and Mesos are most often mentioned. As is generally the case in the FLOSS arena, evolution is a constant. Therefore, if one is in the DevOps arena, it is time to familiarize yourself with containers if you have not already done so.
First of all, let's examine the virtual machine space. VMware, the elephant in the room, still commands a lot of attention. After all, it is a publicly traded company with billions in revenue. A lot of companies utilize VMware for its virtualization needs, and the numbers bear that out. Essentially, it is still the leader in the clubhouse. But, its lead is not what it once was. Over the 15 months examined, whether it is VMware, ESX/ESXi or vSphere, all have been on the decline. Each of those keywords within the listings has seen a 20-30% slide. And, more concerning is the fact that they have been falling at a fairly even pace.
Next up, two prominent projects in the FLOSS arena that perhaps have lost a little of their luster. I am speaking specifically of Xen and KVM. We all watched Xen turn into a professional company, Xensource, and then was acquired by Citrix. KVM followed a similar pattern. What started out as KVM turned into its professional entity, Qumranet, albeit briefly, before it was eaten up by Red Hat. Both Xen and KVM have seen declines during the time period, although Xen is deteriorating at a faster clip. In each of their defense, they were both purchased by large firms, so perhaps the numbers are lost in translation as a result. However, for this particular exercise, we must focus on keywords, and in this case the results are what they are.
Demand for Virtualization Technologies
As some fall, others rise. That is the case with containers. For ease, let's start with Docker. After all, it is considered the pioneer and is the biggest player in the container space. Much like Xen and KVM before it, it too has been developed into its own company, Docker Inc. Too early to tell if it will follow the same fate as those two predecessors. However, we can not worry about that at this point. The more important piece is the traction it has gained. Its rise has been phenomenal. In a short 15 month period, its growth is 3x. It is one of the fastest growing mainstream FLOSS projects we have analyzed. Truly amazing! In fact, it is asked for by nearly half as many employers that ask for VMware. That says quite a lot when you consider the length of time each has been around.
Not to be forgotten is Kubernetes. Originally an in-house Google project, it has since been donated to the Cloud Native Computing Foundation. While Docker's growth rate of 3x seems incredible, and indeed it is, Kubernetes has had a growth rate of 7x. Astonishing! Perhaps the largest that we have seen in all the numbers we have researched.
Finally, Mesos. Apache Mesos is another piece of software in the container space on the rise. While not quite keeping up with the likes of Docker and Kubernetes, it has still commanded a growth rate of 2x over the time period examined. Surely nothing to sneeze at.
Demand for Containerization Technologies
So, where does this leave us? Is the virtual machine space giving way to containers? In essence, yes. The numbers indicate that to be the case. Each of the popular container projects analyzed is currently outpacing its counterparts in the virtual machine space. And, with the Linux Foundation forming the Open Container Initiative (OCI) to establish open industry standards around container formats, its existence is cemented. As a result, is the traditional virtual machine space doomed? At first glance, one might tend to think it is headed in that direction. However, much like in other areas of technology, there is always a lot of dated software that large corporations run that does not translate well to the container space. Therefore, virtual machines are not going away anytime soon.
In an ever-changing world, the data surrounding relational and non-relational databases is no different. While there are proponents for both, it seems to be a case by case basis for which is best for a particular environment. And, with the pace of innovation, the answer can be swayed on a frequent basis. With all that being said, the numbers don't lie, and there is evidence of a shift that is occurring.
While most of the stalwart SQL related databases (MSSQL, Oracle RDBMS, DB2, etc.) remain stagnant, there has been a decline in interest for MySQL. Might seem surprising given the fact that one of the core components of the original LAMP stack is losing ground. How could that be the case? In essence, a lot has changed since it came into prominence. Most importantly, in my opinion, is that it was purchased by Oracle. No longer an independent entity, you have one company controlling two prominent database choices. Is it possible for Oracle to be unbiased and treat each on its own merits? Perhaps. Only the folks inside the company know the true answer to that. Whatever the answer may be, it goes without question that the momentum that MySQL once had has been subdued.
People that have paid attention to the history of MySQL would automatically assume that the decline in the use of that database must have been transferred to MariaDB. After all, that is the offshoot of MySQL that occurred after Oracle purchased it. But, the numbers are not indicating that to be the case. While the number of job postings indicating an interest in finding someone with MariaDB skills have gained over 100% over the last 15 months, the overall numbers are still rather insignificant in the grand scheme of things. They do not make up for the overall drop in MySQL over that same time period.
Therefore, who made up for the gap? From the numbers being analyzed, the one area that has seen significant gains is NoSQL related databases. These non-relational databases have increased by nearly 25% over the time period. Not surprisingly, it is being led by MongoDB in overall numbers. However, after an initial surge in the first three months of data collected, the results of MongoDB have somewhat stagnated as well. So, where is the growth coming from? The numbers are telling us that it is coming primarily from HBase and Cassandra. Apache HBase and Apache Cassandra have seen growth rates of roughly 50%. More importantly, in both cases, the growth has been steady at each time interval. These two, of the more well known NoSQL databases analyzed, are the only ones that have shown that kind of consistency.
However, even after trying to break down the data to determine where the growth is coming from, challenges still persist. After all, this is analyzing employment listings, and they have a tendency to be vague at times. For instance, here are a couple of bullet points from a more common listing requiring a need for someone with knowledge in NoSQL databases:
Thorough knowledge and deep understanding of internals of NOSQL approach
A keen interest in NoSQL and Cutting Edge technologies
Thus, there are a number of listings that provide a lack of understanding of what particular technology they are using. Is this company utilizing MongoDB, Cassandra or HBase? All are a possibility, as well as another NoSQL derived database. With that being said, at least the result would fall under a general NoSQL category, so we can still confirm the growth trajectory in this area.
At the end of the day, what does all this data tell us? The large SQL related databases keep moving along in a consistent, yet mundane pattern. This is, more than likely, largely due to their installation base being large corporations that have the tendency to move at a more methodical pace. The exception being MySQL, which has seen a drop over the time period evaluated. And, of course, to their slow demise comes the rise of the NoSQL databases. In simplistic terms, there is a transition that is occurring from relational to non-relational databases. Probably not too surprising with the emphasis on “big data” in today's environment. So, will this momentum continue? I would expect so, but we will only know for sure as we continuously get in new batches of data in the months ahead.
While the overall numbers are not necessarily sexy, the percentage growth surely is. If you are looking for a tool to search and analyze data, Elastic stack and Splunk are showing exponential gains in the last couple of years. Therefore, if you are currently a DevOps engineer or looking to take your career in that direction, it is essential that you familiarize yourself with the Elastic stack or Splunk.
In order to analyze the Elastic stack, we broke it down into its main three components: Elasticsearch, Logstash, and Kibana. While all three separately have grown over 40% over the course of the last year, it is essential to put the trio together to get the full extent of their growth trajectory. In that case, we have seen a growth rate that has exceeded 150% since the latter part of 2015. Truly remarkable.
LinuxCareer.com is an independent web portal examining a wide range of GNU/Linux and FLOSS related affairs.
We specialize in FLOSS based careers and closely related Information Technology fields. Our goal is to provide readers advice on career advancement and inform them about current employment opportunities.
We are not affiliated with any local or international company, nor are we a recruitment or employment agency.
Editor in Chief
We are looking for real world GNU/Linux stories! Send us your story or topic tips:
editor (@) linuxcareer.com
For website issues and difficulties contact:
admin (@) linuxcareer.com
Have a general question for us? We'd be happy to help you find the answer: