Home > Death by a Thousand Cuts - A Driver for Network Automation

Death by a Thousand Cuts - A Driver for Network Automation

At what point does the frequency and volume of “it will only take a second to change” become too much to bear and force us to adopt a network automation strategy? Where is the greatest resistance to change? Is it in the technical investment required, or is it the habit of falling back to the old way of doing things because it's "easier" than the new way?

The Little Things

We all have those little tasks that we can accomplish in a heartbeat. They're the things we've done so many times that the commands required have almost become muscle memory, taking little to no thought to enter. They're the easy part of our jobs, right? Perhaps, but they can also be the most time consuming. There's a reason those commands have become so ingrained. We perform them far more than we should, but haven't necessarily figured that out yet... well, not until now anyway. The solution? Network automation! Let's get all of those mind-numbingly simple day-to-day tasks taken care of by an automation framework so that we can free ourselves up for work that's actually challenging and rewarding. It's that easy! Or is it?

The Huge Amount of Work Required to Avoid a Huge Amount of Work

Automation, even with the best of tools, is a lot of work. That process itself is something that we will wish could be automated before we're done. There's the needs analysis; the evaluation and selection of an automation framework; training of staff to use it; and the building, documentation, and maintenance of the policies themselves. When a significant portion of the drive for automation comes from overload, the additional technical workload of building an automation framework in parallel to the current way of doing things can be daunting. Yes, it's definitely a case of working smarter rather than harder, but that's still hard to swallow when we're buried in the middle of both.

The Cultural Shift

People are creatures of habit, especially when those habits are deeply ingrained. When all of those little manual network changes have reached the point that they can be done without real thought, we can be absolutely sure that they're deeply seated and aren't going to be easy to give up. The actual technical work to transition to network automation was only half of the challenge. Now we have to deal with changing people's thinking on the matter. Here's a place where we really shouldn't serve two masters. If there isn't full commitment to the new process, the investment in it yields diminishing returns. The automation framework can make network operations far more efficient, but not if everyone is resistant to it and is continuing to do things the old way. There needs to either be incentive to adopt the new framework 100% or discouragement from falling into habitual behaviour. This could even represent a longer process than the technical side of things.

What Must Be Done

Neither the technical hurdles nor the human ones remove the ultimate need to automate. The long-term consequences of repeatedly wasting time on simple tasks, both to individuals' technical skills and job satisfaction and the efficiency of the organization, makes a traditional approach to networking unsustainable. This is especially true at any kind of scale. Growth and additional workload only serve to make the problem more apparent and the solution more difficult to implement. Still, there's no question that it needs to be done. The real questions revolve around how best to handle the transition.

The Whisper in the Wires

It's difficult to say what's harder, the technical transition to network automation itself, or ensuring that it becomes the new normal. By the time we reach a point where it becomes necessary, we may have painted ourselves into a corner with a piled-up workload that should have been automated in the first place. It also represents a radical change in how things are done, which is going to produce mixed reactions that have to be factored in.
Jody Lemoine
Network Greasemonkey, Packet Macrame Specialist, Virtual Pneumatic Tube Transport Designer, and Connectivity Nerfherder. The possible titles are too many to count, but they don’t really…
Read more