helpwanted

Popular thought is eventually every product gets commoditized. Some industries may take longer than others, but competition normally catches the market leader and costs a driven down. Or more efficient ways of providing a product or service is developed. Vendors then have to find some way to add value around the commoditized product. I’m wondering if enterprise infrastructure engineer is quickly becoming a commodity and what value-add needs to occur to stay happily employed.

We have heard it preached that computer science is one of the sure career bets. You constantly see a need for computer science majors in the workforce. However, IT engineers are under pressure from several fronts. There are Cloud services such as AWS, VMware Cloud Hybrid Service and the host of OpenStack providers competing for your compute responsibilities. Then there are the traditional outsourcing companies such as Accenture, IBM and HP Enterprise Services competing for general IT services. Both of these trends have been going on for a few years now. Adding to these pressures are the hyper-converged infrastructure and Software Defined Networking (SDN).

Our industry is a double-edged sword. One of the most compelling parts of the industry is how quickly it moves, and the need to constantly re-invent self. But it’s this same innovation that can bite the career. We find ourselves in situations where we become as smart as possible to make our own jobs obsolete. Every year it’s a grind to do more with less. And every year the bar is set even higher.

“Hyper-converged Infrastructures” is an interesting development during the recent past. Enterprise technology is unique in that as it matures, the physical size of the equipment shrinks while the performance and capability increases. This has enabled a new class of datacenter devices that combine shared storage, compute and the connection between the nodes (networking). Examples of companies putting out products in this category include Nutanix and SimpliVity. The advantages from a technology perspective are pretty obvious. You eliminate the cost and complexity while achieving great or good enough performance.

Hyper-converged Infrastructures allow administrators to do more with less. Not just less physical resources but also with less engineering staff to support the environment. For the right installation, converged infrastructures can reduce, eliminate or consolidate entire support pillars. Converged Infrastructures are aimed at reducing complexity and costs. SDN also adds a similar type of pressure.

The network will get virtualized. There maybe a question to what extent and which technology. The enterprise will eventually embrace the network overlay. There will always be a need for engineers to manage the physical layer of the network. Just like when carriers implemented overlays such as MPLS, there was a need to continue maintaining the physical underlay. However, it did enable the carriers to reduce the costs of providing services to more customers using less infrastructure and engineers.

SDN will have the same impact on enterprise. The promise of the technology will allow enterprises to offer better service levels with less infrastructure and fewer engineers. This isn’t a new trend in technology. IT engineers ways to do more with less every day. You should be able to look over your resume and see how you’ve enabled an enterprise to leverage technology to do great things while either reducing cost or at least slowing down the growth of IT spending. In other words, the measure of your success is how well you are able to do more with less each day.

This all means the nature of the jobs change with more frequency than most of us would like at times. If you are an engineer in a pillared organization, you should look beyond your pillar organization to a broader skillset. The individual pillars are being broken down providing an efficient infrastructure. The need to have a deeper understanding across IT infrastructure is critical. But that’s the simple part. As engineers, we understand upgrading our technical skill to meet the market. Just upgrading your technical skill over the long period will help you stay employed in technology. With the direction of the market, the job may be with a cloud provider or service company.

What do you do if you enjoy working in end user organizations? To do so, enterprise engineers need to become more familiar with the business of business. They need to add value around the commodity of enterprise service. Enterprise engineers will be required to take business requirements and translate them into technical requirements. Think of the role as a management consultant with a focus on enterprise IT. Today it’s a unicorn of sorts. The team over on Geek Whispers talked about this mystical beast – seasoned IT worker with deep business skill.

I may be biased. My day job is an IT Management Consultant. I get to speak geek and business every day and see the future of the enterprise IT worker consolidating. What are your thoughts? Can you stay a pure geek and thrive in corporate IT long term?

The future of the enterprise engineer will be more than technical
Tagged on:

19 thoughts on “The future of the enterprise engineer will be more than technical

  • January 20, 2014 at 11:48 pm
    Permalink

    I think far too often we think technology for every company moves a lot faster than it does. Case in point is that I’m sure that there is a large number of companies that have a few or no virtual servers in their environment. BYOD is supposed to be the new user access paradigm yet there continue to be companies with thousands of PCs. VDI was supposed to kill the PC and EUC is supposed to kill VDI yet neither has been very effective in doing much killing. Unfortunately it is far to cost prohibitive to try the flavor of the month (SDN, SDS, etc.) without an appropriate use case for their company. While automation is one of the initiatives in IT right now it is far too difficult to implement for most environments to make it worthwhile. The complexity of the task often requires a programmer and then an engineer to tell the programmer what the task is. Why would I pay two people to do the task of one person. If there are still people who work on mainframes I feel confident that it will be quite sometime before we’re replaced by robots.

    Reply
    • January 21, 2014 at 2:47 am
      Permalink

      Sure there will be paid exceptions but your mainframe example is perfect. They in large have been replaced. The industry needs far less personnel to maintain them today. The question is how do you become one of the few.

      Reply
  • January 21, 2014 at 12:12 am
    Permalink

    I am seeing the same thing, You bring more value if you can interpret business requirements into IT solutions and understand the domains that will support them. Top performers are being asked to understand compute, storage, the hypervisor, the network, as well as the political and financial layers. My Director told me that he hired me due to the thought leadership and vision I could bring to the business, as well as my technical skills, but mentioned he could contract out someone to implement my design. You can get on the train when it stops, you can run after the train when it starts to leave, but when the train is gone…The roles of a Senior, Lead and Architect are being re-defined as well. More is being expected, and this is the way we have always done it mentality is going South.

    Reply
    • January 21, 2014 at 9:29 am
      Permalink

      That’s a pretty good example of what I’m talking about. We will see the shifting continue to happen of the next few years. Highly technical people will work for integrators and providers while the more rounded SEs will work in the enterprise.

      Reply
  • January 21, 2014 at 2:15 am
    Permalink

    In my country/language, the Netherlands/Dutch, the field of IT used to also be called: automation.

    A clear reminder of what we are really doing.

    So if you think about it, if that is the purpose of our field: that would mean if you are doing it right, after you’ve automated something, you shouldn’t actually have to touch it again any time soon. And you move on to automating something else, other projects.

    Reply
      • January 26, 2014 at 9:38 am
        Permalink

        Had an other thought. So the answer probably is: move up the stack.

        That is how it has always been. Most programming isn’t done in assembly language these days. For example less and less programmers manage their programs memory. Most of the programmers are doing scripting these days.

        For example maybe you are at a company that isn’t using SaaS so much, but uses IaaS, instead of bare metal in your own datacenter or using some kind of converged infrastructure (the other Unicorn !). Then there is a lot of stuff on top of that that still needs to be managed.

        Reply
  • January 21, 2014 at 10:07 am
    Permalink

    I speak about this often when I discuss collapsing silos in organizations and the fact that if you are just the “Storage Guy” your days are probably numbered. I’m not saying there is not a need for specialization, or specific focus, but your technical horizon needs to be expanded, because the future of many technologies is to abstract and collapse the complexity.

    At least thats how I view the path forward.

    Reply
    • January 21, 2014 at 10:10 am
      Permalink

      We are on the same page. You triggered a thought. Today that abstraction is based on virtualization software but virtualization is just one method. Your comment I think will lead to another post.

      Reply
  • January 21, 2014 at 3:07 pm
    Permalink

    No, you cannot stay a pure geek and thrive in corporate IT long term. But, you never have been able to! If you want to thrive, you need to know your business. That is step 1 in converting IT jobs to an IT career. It does not stop you from changing jobs, either, but it does require you to learn about the business of each of those jobs. Over the past 15 years, I have, in no particular order, learned far too much about licensed and unlicensed wireless spectrums, registering convicts as they enter and leave the penal system, how to attract and retain clients, how to use AutoCAD and plotters, the ins and outs of renting units in an apartment complex, electrical and HVAC systems, how difficult it is to run a county election, what HIPAA really means in a nursing home, and what it takes to fight forest fires and keep everyone alive. I did not learn these things because they were part of IT or my job (although there’s overlap, particularly with wifi/elec/HVAC) but because that’s what the entities I worked for or with did.

    I think the most important way in which this changes in the future is that abstraction allows us to move closer to being a jack of all trades and master of none with less penalty. Take, for instance, performing backups in the ’90’s. Even if you knew what you were doing, installing a new backup system for an enterprise was a minimum of 2-3 days, possibly weeks, and involved a lot of assembly, learning, and training. In 2014, it takes 30 minutes plus however long it took for you to download the software, and there’s little to no assembly required. Literally, none! When I installed Veeam a few days ago, I saw all the components that were missing from my machine and groaned. In the days of yesteryear, I would have had to install every one myself, then launch the installer and hope that it recognized everything as complete or start over until I got it right, THEN I might be able to get it installed. Instead, there was a button to install all the missing requisites for me. I didn’t even have to acknowledge another license or EULA, I just had to wait a little longer.

    Compare the backup situation – a technology 30 years in the making and refined, refined, and refined until we finally have usable products – to cloud management. vCAC is a great product, but an absolute PITA to install and get running. I have no idea how long it will take, but apparently at least another few years before this gets to be refined to a next->next->next install. This means that right now, if you are good at vCAC, you have to get your fingers in real deep. In the future, you won’t have to be quite so deep in it. You’ll be able to branch out into whatever the current “assembly required” technology is. And that’s a good thing – it means solved problems stay solved and your engagement stays high as you continually move to new problems.

    So, don’t be worried about all the changes. Learn how to ride the wave and enjoy it. It’s only a double edged sword if you try and wrap your hands around it and stop it from moving without you.

    Reply
    • January 21, 2014 at 3:46 pm
      Permalink

      Rob, great points. I think it works both ways in today’s climate. Just like there was a penalty for being too shallow in a specific area back in the 90’s and early 00’s there’s a penalty for being too deep in a specific area.

      To Lennie’s point this is no different from any other industry. There was a time where I could focus on just being the best forklift driver in the plant. Now, no one needs a dedicated forklift driver.

      Abstraction is forcing the majority of us to move beyond the specialist title or else.

      Reply
      • January 21, 2014 at 7:22 pm
        Permalink

        I think it is possible to be a specialist. But to flourish there, you need to know your business. You can be a purely storage guy – as long as that’s an important thing to your business, and if you can use storage to make the business better. It might mean that you have to work for someone like NetApp or Pernix, or work in the bowels of something like AWS or Azure, but there’s plenty of jobs. The business expertise is what moves you from being a replaceable cog to a valued member of the business.

        Reply
  • January 22, 2014 at 10:19 am
    Permalink

    It’s funny the different trains of thought on this. People are constantly saying that you need to specialize but when I see how technology is converging I am left wondering if that’s actually good advice. Look at DevOps. Try telling those guys that all you need to know is server administration. If you are working in a culture of DevOps, using on an on premise cloud but bursting out to AWS, how can you not be required to know all the moving parts. Those guys/gals have to know security, sys admin, networking.

    Reply
    • January 22, 2014 at 10:21 am
      Permalink

      Good example. I do see the need for specialists. I just don’t see them long term in end user orgs. As Rob Nelson pointed out in his last comment, companies like Amazon, Google and Rackspace will always needs specialists as they provide services to end user organizations.

      Reply
    • January 23, 2014 at 5:24 pm
      Permalink

      My experience is that the top people at DevOps shops know the moving parts well. Everyone else knows them well-enough. I would have to say that most of those people are still fairly specialized and shallow in knowledge of other areas. That’s not a problem, except when organizations think that many people’s shallow experience is equal to one person’s depth of experience – one good security engineer is worth a dozen security hacks. All you need to do is look at a list of the latest security incidents to see what “well-enough” gets you in that realm.

      Reply
  • Pingback:The year of the Unicorn – TechTalk 34 – Virtualized Geek

  • January 24, 2014 at 9:38 am
    Permalink

    Keith, I currently work as an IT generalist/ Enterprise Administrator and at my company we are working with everything Microsoft has to offer. I have heard of the big shift that will take place. Basically, two guys will manage the data-center and the Enterprise Admin will need to re-tool their skills. I thinking that I may need to specializing in the System Center Suite and become a consultant but after watching your video that may not be the best thing to do.

    Reply

Leave a Reply

%d bloggers like this: