Network Automation Has Been Elusive; What's Holding Us Back?

Summary

For decades, network automation has stood as a pivotal ambition within the networking industry. Yet, despite literally millions of person-hours invested, more than half of companies have yet to succeed. Even among those who have had successful automation projects, NetBox Labs reveals that a staggering 57% of network tasks remain manual. Essentially, the networking industry remains stuck at approximately 30% automation, despite massive effort.

Why have we fallen short?

Before we can collectively plot a path forward, it is crucial to understand why network automation efforts have largely fallen short. Is it because we're waiting for a breakthrough tool to revolutionize the industry? Or is there something more fundamental at play? Perhaps the answer lies not in what we're developing, but why we're developing it.

Talk to any IT executive, and they'll tell you that agility and speed are top priorities. But how can you automate anything when just keeping the network running requires herculean efforts?

The truth is, you can't. And yet, the networking industry is full of companies promising to deliver agility and speed through their products. We see this in the proliferation of workflow engines and other tools designed to build bigger automation engines.

What if our industry's challenges in solving for operations are rooted in something deeper? What if we've been pursuing the wrong "why" all along?

Let's ask a question: If you had a tool that could push all your team's proposed changes into production without any additional effort, would you use it? The answer is a resounding no. We know that when we make changes, our fragile networks don't always survive. Automation may reduce effort, but it doesn't ensure that our networks actually work.

The real challenge is reliability.

Anyone experienced in automation will tell you that it's the fastest way to break things at scale. How many outages out there were caused by small human mistakes amplified by automation systems? Too many. I'm not anti-automation, quite the contrary, but I believe that the core challenge to solve isn't speed–it's reliability.

We need to eradicate human error. If we can guarantee that proposed changes will work, we can move quickly and confidently. If our tools emphasize correctness and repeatability, we'll reap the benefits we've been after all along.

Here is a new approach.

It's not that speed isn't important, but the path starts with a different purpose. By driving human error to zero, we'll achieve speed and efficiency as byproducts. Predictable automation is key–delivering reliability across data center networks and networking.

With tools that combine reliable outcomes with  ease of use, you can finally eliminate the fear factor around changes. By incorporating features like revision control, pre- and post-checks, digital twin, network-wide transactions, and more, you can have confidence that your network always operates the way you want it to.

So, what does this look like in real-world scenarios?

Imagine a network where changes are validated and verified before they're implemented. Where automation tools don't just execute workflows but guarantee correct outcomes and emphasize repeatability. This is the future of network automation–one where reliability and speed go hand-in-hand.

In the end, network automation is a complex challenge, but by focusing on reliability and eradicating human error, we can achieve the speed and efficiency we've been striving for. It's time to rethink our approach and prioritize the right "why." With the right mindset and tools, we can finally achieve the elusive goal of network automation.