The San Mateo, California-based company was founded in April last year as a result of conversations its CEO Allan Leinwand had held with IT staffers in the financial services industry.
They were already using open source in their servers to migrate away from high-end, proprietary Sun boxes, which brought them to the next line in their budgets: networking, said Dave Roberts, Vyatta’s VP of strategy and marketing. They’d cobbled together a mishmash of packages as proof of concept, but there was no-one to stand behind the project with a standard distribution, services and support.
Vyatta was formed to fill that void. It is VC-funded, with a next round of financing to be announced within the next couple of months, and sees services and support as its sources of revenue going forward.
If that sounds distressingly similar to the plans of many a Linux start-up back in the dot-com bubble, Vyatta can always hold up Red Hat, and even SuSE prior to its acquisition by SCO, as success stories in open source.
In any case, Roberts said the main focus for Vyatta is in building the developer community, not commercializing a product, and so it has made a beta, or what he called a 0.5 version, available for download on its website at www.vyatta.com. Version 1.0 is expected in the Q2/Q3 timeframe, he said.
Roberts defined two market spaces at which Vyatta’s Open Flexible Router (OFR) is aimed. Firstly there is the branch office with a T1/E1, T3/E3 or OC3 connection to a WAN, and secondly, on the corporate LAN backbone, delivering data rates of between 100Mbps and GbE.
The product is not for the high-end, carrier class, he acknowledged, which is where Cisco and Juniper have highly tuned, robust offerings, but look at how Linux has gone from one-way X.86 configurations to 256-way SNMP environments, he said.
Content from our partners
As for the software Vyatta uses to build OFR, Roberts said the product is Debian-ish in the baseline packages, with lots of optimized networking-specific additions.
These include its support for mainstream routing protocols (BGP4, OSPF2 and RIP2, as well as static routing), which comes from the XORP project for open source routing at the International Computer Science Institute at Berkeley, which Vyatta now sponsors.
While OFR is currently a software-only offering that can be downloaded and blasted onto a standard server, it would come as no surprise if Vyatta at some point launched an appliance to run it on, not least because such hardware could also provide a platform for other services besides routing.
Roberts wouldn’t be drawn beyond acknowledging that a dedicated hardware component would represent a very logical step.
As to those other services, he commented that today we’re focused on routing protocols and firewalling, so VPN is a natural add-on, and we’re being inundated with requests to support WiFi.
Roberts cited Linux’s progress from desktop to mainframe as an example of how far an open source platform can go, given the right momentum. To create that momentum, however, Linux required the endorsement of a heavy hitter in enterprise IT, namely IBM, which embraced Linux wholeheartedly from 1999 on, taking it beyond the realm of the ponytailed geek into mainstream enterprise computing.
For the same thing to happen to OFR, it will need a similar patron. Roberts said Vyatta had already received some early shows of interest, but it’s still early days. Who would be likely candidates for the role? Certainly not Cisco or Juniper, which have invested too much in their proprietary OSes to see their user base cannibalized by an open source alternative.
Perhaps a challenger like Nortel of 3Com, each trying to get back into the game, or one of the newer players out of Asia (e.g. Huawei or ZTE from China, Allied Telesyn from Japan, D-Link from Taiwan) might care to step up to the plate. Or perhaps even Dell, the great commoditizer, which despite a couple of forays has so far failed to replicate in networking the impact it has had in desktops, servers and now printers.
