NFD27 - IP Fabric

Day two of NFD27 kicked off with IP Fabric. I hadn’t heard of them before, but I’m in North America and they’re in Europe, so I guess that’s a bit like most North American operators not being familiar with Nokia (f.k.a Alcatel-Lucent). We’re going to see a lot of overlap with Forward Networks, the company that wrapped up the first day of NFD27 (and was previously covered here).

First Impressions

I mentioned there’s some overlap with Forward Networks, and is there ever! Based on the presentations, I think they’re competitors. Forward is based in North America, whereas IP Farbic is in Europe. I think this is an important distinction to make that is sometimes ignored: it’s hard to sell when your timezones aren’t aligned and when you have language and cultural differences. Global sales are hard.

But let’s stop talking about trying to manage a global sales pipeline and start talking about IP Fabric!

Let’s get the buzzword out of the way: Network Assurance. That’s their pitch, and even if Forward doesn’t use that phrase, that’s what Forward is doing, too. A lot of companies and/or products are throwing that phrase around, too: Juniper uses the phrase when referring to its Mist platform. But what’s special about IP Fabric?

Perhaps most importantly, they support a wide range of vendors and technologies. More than just traditional on-premises network devices, IP Fabric also supports public cloud. It leverages the APIs of those providers and can understand network-related items such as VPCs, firewall rules, Transit Gateway (for AWS), and more. It can then combine that understanding with its understanding of on-premises networks to paint a picture of end-to-end hybrid network flows.

Like I mentioned in my thoughts on Forward, this is a powerful capability. As an operator of a hybrid network, I would love to have the ability to look into the path a packet takes from my on-premises network all the way into the public cloud. Perhaps more important than that, though: I want to share that insight with other people at my company. I want to enable and empower developers, support teams, and server/systems teams to understand what’s happening when things go wrong or how something works.

Final Thoughts

The interface was intuitive, but the overall maturity was somewhat lacking for my use cases. Specifically, while they do support public cloud providers, their support for Google Cloud Platform is not as mature as their support for AWS. There’s also a lack of Kubernetes support. Maybe that’s not important to everyone, but if you’re in an organization leveraging Kubernetes, this is an pretty major feature that could be game-changing, especially for anyone new to or unfamiliar with the internal networking.

This isn’t really a dig at them, though. The nature of product development is that you’ll deliver some features before others, and I’d rather have that than wait 5 years to get everything.

Published At
Tagged with