Stanford team behind NSX and OpenFlow spawns network modelling startup

The Stanford University research team that gave the world OpenFlow and VMware's NSX has spawned another interesting startup: Forward Networks de-cloaked on Monday US time with a network modelling technology.

The company's tools let you take a snapshot of your networks and then build a working simulation in software. “Think traceroute, but without sending packets, and with the full layer 2 through 4 specifics of what happens at every device along the path,” explains CEO and co-founder Co-founder David Erickson, who reckons this will be handy because today it is hard to predict all possible packet flows. Forward aims to offer the chance to do just that and then to analyse traffic to see if it conforms to policy so you can take corrective action if packets are heading in difficult directions.

A third piece allows modelling of impact network changes, the better to predict any unforeseen consequences that flow from changes to network configurations or the addition of new kit. Or just to mess with your network without having to buy more kit, in the hope you can get a bit DevOps-y with your network on a search for continuous improvement.

This is a big end of town play: Forward's plan is to win telcos, financial services outfits and IT services providers as customers and can already name Australia's dominant carrier Telstra which is apparently using its kit “to increase operational efficiency of some of their key clients’ networks.”

“You really had to have a breakthrough in Computer Science algorithms to do this,” Erickson told your correspondent.

Forward's algorithms are the result of many years effort and have their roots in research conducted at The McKeown Group, Stanford's team dedicated to next-generation networking problems. During his time with the group Erickson was mentored by NSX daddy Martin Casado, who now works at venture capital outfit Andreessen Horowitz, which has funded Forward Networks.®