More and more organizations are weighing up the numbers – is it still cost-effective to run your own data center or even rent rack space in someone else’s data center? Is Infrastructure as a Service the way to go, to smooth out periods of large capex and reduce the hidden costs of hardware maintenance?

Your results may vary, whether you decide to keep it all in house, just use IaaS for projects or decide to decommission all of your own tin.

IT Professionals across the spectrum of roles are looking at how the industry they operate in is shifting. They want to know how they can keep up with the demands of a business that can easily provision IT services on a corporate credit card. Even without a shift to Cloud, concepts like Infrastructure as Code and Software Defined Networking means your own data center may be a very different place than it was 5 years ago (heck, even 2 years ago).

This was highlighted in a recent Ask the Experts session at Interop 2016 featuring Chris Wahl, Jason Nash and Scott Lowe.

Infrastructure engineers are showing concern about this move to the Cloud and where they fit into the equation. While application developers may still have their apps to patch and upgrade, what’s left for the infrastructure engineer if you take away their infrastructure?

The experts recommend that you “de-couple” yourself from the notion that infrastructure equals hardware. Apart from the physically plugging something in, replacing a failed part or adding an upgrade, the infrastructure engineers’ role extends to managing and monitoring the infrastructure environment and the performance of that hardware, as well as extending and integrating the infrastructure. That ‘Infrastructure Ops’ and architecture skills are where you want to focus your attention.

If you’re keeping your own servers, what advances in infrastructure management can you take advantage of and how will that ultimately benefit the business? It’s time to move away from snowflakes and start to automate and duplicate your configuration standards. If you’ve got more than one server performing the same role, they should both be viewed as ‘brown cows’, not as Daisy and Clarabelle.

But with options like Chef, Puppet and Ansible, what should you use?


The answer here is “it doesn’t really matter”. Organizations often don’t move forward, because there are too many choices and they are scared of betting on the “wrong” tool. The reality is there isn’t really a “wrong” tool, just like there isn’t a perfect tool. It’s not the tool you choose that will ultimately make the biggest difference – it’s how well you use it.

These skills will translate to Cloud-based IaaS too. If you’re provisioning IaaS hardware using a GUI every time, you’re doing it wrong.

But the skill that will have the biggest impact for a business adopting IaaS is accurate sizing of infrastructure resources, both during provisioning and as the use of the service settles down into a normal pattern. It’s natural to want to size infrastructure with room for growth especially if you have no relevant historical usage or performance data. With IaaS, you’re paying for that extra capacity every month, so why keep it if you don’t need it? The Cloud provider isn’t going to tell you that your infrastructure is over-sized. Sometimes the business doesn’t tell you that a project is over or has been drastically reduced. Learn how to keep an eye on the workloads on an ongoing basis to see where you can release IaaS resources and save the business money.

And remember, it’s still a good idea to monitor IaaS availability (or the availability of the system it is running), to the extent that you can, so you’re not relying on the Cloud provider to tell you there’s an outage.

Change is constant in our industry, but that doesn’t mean it’s always easy to adapt to and embrace. Let us know what’s changing in your infrastructure world and how you’re embracing new tools and skills.