There is no doubt that virtualization technology has been one of the leading factors in the dramatic changes we have seen inside the 21st century data center. For all practical purposes, the landscape in data centers today looks a lot different than it did before the turn of the century. During this evolution that fundamentally changed the technology environment, how much has the support structure changed to keep up with virtualization and cloud technologies? This is the topic that I would like to focus on in this multipart post, which will explore the different support structures and concepts that have been the standards.
Let’s start a discussion about what just might be a better way of doing things moving forward. First, I want to establish the baseline before moving on to the different thought processes and philosophies for the future.
One of the main factors that can define a company’s support model is, quite frankly, the size of the infrastructure and the company itself. The small and medium-sized companies typically have fewer resources and less money in the budget for IT. As such, these smaller companies are always looking to get the most return on investment for each and every purchase that is made. The support staff are usually “do it all” types in that they are expected to support and maintain all of the different technologies in the environment. These technologists are truly the masters of their own domain. Deviations from the default when it comes to deployments and operations are few and far between. Adding complexity to the environment is not something that is done too often, unless it is truly needed to solve a specific use case in the environment. Further, these “masters of all” administrators tend to have a good basic understanding of how each of the different technologies works, but they are missing the real technical depth needed to deviate very far from the default or the norm.
In what could be considered a direct correlation with the size of the staff and the infrastructure, as we look at medium to large companies, we start to see the establishment of technology silos and, with that, technology subject matter experts, or SMEs. At first, each of these silos specializes in certain technologies, and team members cross-train with other team members to be secondaries or backups to another of the technology silos. This is the start of the separation of skill sets. Teams are able to focus on specific technologies and, as a result, develop a deeper understanding and sharper skills for their subject matter.
In a natural progression, as technology departments and infrastructures increase in size and scope, there is an increase in the number of different silos and a decrease in the amount of cross-training into different technologies, until you get to the largest environments, in which the silos are seasoned administrators with a very in-depth, in-the-weeds understanding of their specific technology areas. With this increased knowledge and understanding comes the best overall capacity for development and customization of the technology. These are the technologists who don’t just work with a product or technology: they are the people who have the ability to truly make the technology their own.
During my career, I have had the opportunity to work in all of these different types of support scenarios and have been what I would like to call “the expected standard of support.” The main point I would like to make with this observation is that the larger the environment gets, the more chances for specific customization and deviation from the default installation and deployment. As the number of supported assets increase, so will the use of scripts and automated processes. When you have a few servers to support and maintain, manually clicking buttons in the graphical interface to perform the support operations works out just fine, but when you get to a point where you are supporting thousands of servers, automation is only real way to maintain and support the environment with any kind of consistency throughout it.
If you have a different understanding or expectation regarding support in the 21st century data center, please post your thoughts in the comments section to help define the baseline. My next post will start with this baseline, look at what works well and where there could be improvements, and present some thoughts about what changes might be needed to make things better and more efficient for support teams of all different sizes.