Q&A: Acra discusses his plans for Procket Networks

Former Cisco Vice President Roland Acra takes the reins at a router start-up.

Six months after losing its CEO, router start-up Procket Networks has landed a new one: Roland Acra, formerly Cisco's senior vice president and Service Provider CTO. Acra, who reported to Cisco Chief Development Officer Mario Mazzola, jumps to Procket as Cisco keeps the industry waiting for its next-generation high-end router and operating system. On his first day at his new job, Acra shared some thoughts with Network World Managing Editor Jim Duffy.

What attracted you to Procket?

The team, first and foremost. It's a great bunch of people, a very, very strong networking team.

A great technology, great products that are out there shipping. All of the technology risk was now removed from the product despite the very cutting edge nature of what has gone into the product in terms of silicon innovation and next generation operating system structures.

And then my feeling that the timing was right. I think over the next 12 to 18 months there will be an upswing in the market for high-end routers. So being there now and building the sales momentum and focusing on the sales effort is good timing.

So is building up the sales and marketing momentum your first priority, or is it on the next generation of development?

Procket has a very strong technology edge to build on right now in terms of extra capacity or density, performance levels, the silicon capabilities and the software architecture, that now need to be exploited outbound in terms of sales and marketing. Definitely the priority will be cranking up the revenue engine and establishing both technology preference directly with customers, and then complementing that with partnerships where appropriate.

Does that plan include engagement with incumbent vendors?

There are discussions going on with partners when it comes to market coverage capabilities and these are all underway. There's nothing there to be announced yet. It's part of the go-to-market approach, to work on partnerships.

[They include] a pretty broad spectrum of domestic and international [candidates] from the telecom industry, either transmission or general data-over-voice communications folks, for resale.

How did you view Procket while at Cisco?

I had the view that they had, by far, the best available technology on the market today. So when it came to technology edge -- in terms of performance, density, structural reliability, in terms of what software and hardware architecture underlies the product -- the level of silicon integration and innovation is absolutely unique. This is not a feature which somebody else can do in 60 days and say, 'me too, I've got it now.'

Just about everyone I have talked to [at Procket] is the best that there is in the industry -- the software folks, the hardware folks, the architects.

Is Procket doing anything technologically that Cisco should emulate?

Procket is building its product on a few key premises, which I think are going to be the definition of what next generation routers are about: one is a very advanced level of silicon integration, and a combination of speed and feature capability. The integration is important because through the integration into custom chips, you get a lot of savings passed on to the customer. You get a lower price-point because of your cost of goods reduction; you get lower power dissipation, you get better density and space usage; and then the ability to have feature programmability at the same time as wire-speed performance means that you're protected against any new demand your network will place on you in the future.

On the software side, where Procket has written the book is in developing the most modular and cleanly designed operating system . The compartmentalization of the different protocols makes for not only a very reliable system, but also for the best cycles of time to bug fix, add   new functionality,  or reduce downtime every time you're considering a planned event on the network -- such as an upgrade -- or an unplanned event, such as a bug that you need to address.

What are Procket's challenges?

Procket has to go through what many start-ups have to go through: you build on a base of technology edge that you translate into benefits to your customers. You strive to make sure that these are not navel-looking but much more 'This is why it matters to you, Mr. Customer, in terms of economics, reliability, future-proof investments.' And then you work on concerns a customer might have on working with a start-up -- the fact that you have a strategy that's credible, that you're prudent about how you manage your product development and finances so you're going to be around for a long time, the fact that you're building a great company as opposed to something completely near term and opportunistic. And you begin by addressing the innovators on the customer side, the thought leaders who have had a history or track record of taking a risk on a start-up and making them viable in terms of references, etc. And then you work your way back towards the more incumbent folks who are sometimes a bit more risk-averse and want to see the other guys prove out the early technology.

The good news is that not only have both products from Procket been shipping for the best part of this past calendar year; but it's also that the technology has been soaking in large customer networks for a lot longer than this. Procket has such strong connection in the large ISP marketplace that they got the investment of time and energy on the part of customers to remove the risk and debug issues in the product for upwards of three years now. There should not be major surprises -- negative ones, that is -- that haven't already been uncovered and mitigated.

Did any frustration at Cisco with either the pace of technology development or organizational structure prompt you to leave?

No, my leaving was entirely a pull model as opposed to a push. I didn't leave as much as I came to Procket.

It has been reported that one of the challenges internally at Procket is in working with the founders, including Chief Scientist Tony Li. How  closely did you work with Li when the two of you were at Cisco?

I worked with him a lot back when I was with Cisco in Europe. I was building the technical field organization at the time. So I was facing the customers, designing IP backbones and BGP routing designs, and Tony would be one of the guys who would be, despite nine hours of difference, on his e-mail answering my questions and helping me through it. We stayed great friends since then and he was, by the way, one of the initial folks who felt me out and approached me about Procket to see if I was interested. I'm very excited to be back working with him.

Insider Tip: 12 easy ways to tune your Wi-Fi network
Join the discussion
Be the first to comment on this article. Our Commenting Policies