Moving Networks Forward
I am thrilled to be joining the team at Forward Networks to take their product to market. The company was founded by a team of Stanford PhDs with an incredibly rich background in enterprise networking. Under professor and Nicira co-founder Nick McKeown, the founders’ work at Stanford focused on building the foundational components of what we now know as software defined networks (SDN). They actually wrote much of the SDN controller software that is widely used by the world’s largest networking vendors.
With funding by Andreessen Horowitz, the founders have since assembled a world-class team of engineers and product talent from Google, Facebook, LinkedIn, Cloudera, and Cisco. Collectively, they’ve spent the last two and half years building a software platform that completely rethinks network operations.
Our mission? Simple yet massively audacious: to help enterprises avert costly network failures stemming from misconfiguration and other human error. Since we’re still in stealth mode, I can’t yet get into any detail on the product itself. But allow me to explain why our mission is such an important one.
Having spent the last half-decade at Meraki and ThousandEyes, I’ve had the opportunity to work closely with the network architects, engineers, and operators of some of the world’s largest enterprises. This experience validated for me what most in the IT industry would readily admit with a groan and a shrug — the enterprise network is a complex, brittle beast.
Networks are easy to break and hard to fix. When something goes wrong and the network falls over, the business impact can be devastating. Just ask United Airlines, the New York Stock Exchange, Facebook, GitHub, Bloomberg, and Telstra — each had a widely publicized data center network failure in the last year. Regrettably, these outages often stem from a rather avoidable cause: human error.
Why?
The network of a typical Fortune 500 company is comprised of thousands if not tens of thousands of hardware devices; including routers, switches, firewalls, and load balancers. Underlying these physical devices, there are thousands or, in some cases, millions of forwarding rules that govern the network’s behavior. This results in a level of complexity well beyond what is easily manageable by even the most adept network operations team.
Operators have little if any visibility into how the simplest of changes in their network will impact the behavior of their devices. A routine change to a Cisco core router, for example, can have unexpected and dire consequences to the traffic flow and behavior of devices downstream. United Airlines learned this the hard way last year.
On July 8th 2015, the world’s 2nd largest airline grounded its flights worldwide for about two hours, which had a ripple effect on flight delays for more than a day. The grounding was a result of a misconfiguration in a router that blackholed the United data center that was hosting the airline’s global passenger check-in application.
Thousands of passengers were justifiably enraged by the grounding. In fact, United Continental Holdings’ shares tumbled 2.7% directly in response to this incident; wiping out nearly $500M in market value in a single day. Network failures just like like this will continue to lead to mass customer defections and damage the profits of any enterprise; be it an airline, a bank, a telco, or a retailer.
Human error is a simple and inevitable fact of life in the enterprise network. Here at Forward Networks, we’re building a game-changing solution with the goal of transforming the way operators and engineers build and maintain their networks. With our software, we aim to reduce the human error, unpredicted behaviors, and configuration issues that plague networks and lead to costly downtime.
So how do we do all this specifically? Well, let’s save that for when we officially launch later this year. Until then, go Forward!
