I was at Cisco Live 2015 this year and had the opportunity to learn about Cisco’s APIC-EM platform, or Application Policy Infrastructure Controller Enterprise Module. In short, it helps manage end-to-end Quality of Service (QoS) and perform-per-flow QoS, in addition to a few other jobs. Ultimately though, it’s another venture by automation software to address what we as engineers struggle with: Can we, as humans, consistently manage some portion of our operating environment when implementation of a feature is inconsistent across equipment?

Why Even Try?

Once I heard about Cisco APIC-EM, I started to hit up some of CDW’s best and brightest to get their thoughts. The questions were good ones. For instance, “How well does it scale?” Also, “If a feature is implemented right, why do you need it?” Moreover, “What are the use cases?”

What I love most about the people I work with is that their inquisitive minds were asking the right questions to know more.

An outside observer would look at these questions and ask, “Why even try?”

My colleagues’ questions also tugged at the primary functionality and elements of the product. Frankly, this piqued my curiosity, too. If everyone who had been skeptical had been right, most of the technology we have today probably would not be here; and automation is no different with its shares of skeptics. But once you start speeding up what used to be slow and labor-intensive, it can be liberating. The key questions our subject matter experts were asking amount to the pillars that need to be strong for it to work at-scale: Maybe APIC-EM is ready…or maybe it isn’t.

What Do You Have To Lose? 

Facebook, Twitter and email are filled with one-liners, deep thoughts, haikus and other literary elements to make us look at mistakes or failure as opportunities. The questions that were asked of the product make me think that this is something we should kick the tires on. We can learn a lot from it, help recommend improvements, fill in the gaps and, with a bit of rinse-and-repeat work, we can end static QoS management.

QoS only works until something does not support a new configuration, or when someone accidentally changes an Access Control List (ACL) and breaks a hop or endpoint. Software can stand to be meticulous and much less error-prone as it is refined.

It is this refining process that makes a strong case for automation. There are lots of things that are just not needed like they used to be. For example, if you build a lot of servers, you eventually flip to Preboot eXecution Environment (PXE) builds. Stateless computers make for better stamped-out systems and remove the “every server is special” mentality. This is the breakneck pace of IT because twisting a knob does not necessarily generate value by itself. Software turning lots of knobs for the right reasons and with error checking can be as thorough as its human counterparts – where humans cannot match the pace.

Getting to a good pace takes some pain, some failure and a lot of learning. As engineers have become experts in deploying technologies and watching minefields, companies will learn what to watch out for and expect from automation. You can build the experience now or later; the choice, expense and timing are yours to make.

“I’d Like My Frontier with Fries, Please”

Today’s out-of-box automation menu is limited and in five years, we will be asking why we ever thought we needed to do what we used to by-hand. By learning to take reasonable steps into automation, you will be learning the tools of the automation trade.

If QoS is on-deck to be automated, what other elements can be automated? Surely, you want WAN optimization, load balancing, AD changes and other features…or fries with that? If you are experimenting now, you will get the skills to sniff out other areas you can automate and draw from what you’ve already learned.

There will always be risks and we know there are lots of rewards to be had…so, why not?

Interested to learn more about Cisco’s APIC-EM platform? Check out this short video to learn more about its benefits, as well as how it can give your IT team the tools they need to overcome network complexity.

Questions? Contact your account executive or leave a comment below.