Q&A: Peeking into Juniper's future

* Q&A with Juniper CEO Scott Kriens

Juniper Networks is one of the few vendors that have figured out how to stand up to Cisco. In core Internet routing, Juniper holds about a 40% market share to Cisco's 55%. Last year, the company took aim at Cisco in enterprise routing, introducing its J-Series WAN routers. Last month, Juniper acquired two traffic-acceleration start-ups, fueling speculation that it might be hungry for even more enterprise network technologies. CEO Scott Kriens recently spoke with Network World Senior Editor Phil Hochmuth about Juniper's enterprise network strategy. Here's an edited transcript:

NW: You've successfully challenged Cisco in Internet routing; what's your plan for gaining share in enterprise routing, where Cisco is more dominant?

Kriens: The strategy is to focus, execute and build trusted relationships with customers for whom the network is critical. It's exactly the same way we grew to a 40% market share in the service provider business from zero. It looked a lot more daunting as a company of size zero to grow to 40% in mission-critical backbone networking. We don't take any success for granted. It's going to be hard work. But we have a formula and model that we're very comfortable with.

NW: What is it about a Juniper enterprise router that would make a Cisco customer want to switch?

Kriens: Before devolving into all the acronyms and gobbledygook, it's this: deep inspection of traffic at high speeds. The compromise in any implementation that we compete against, whether it's firewalls or routers or anything else, is other products can look deeply and carefully at the traffic flow, as long as it's going slow enough. Or they whip it through fast enough to meet the requirement, as long as they don't look very close. But of course the problem is you have to look very deep - closer than ever - inside packets. And as we start to do things such as run video and live TV on cell phones, you have to do it faster than ever.

NW: Regarding your recent acquisitions of Peribit and RedLine, how will their technologies be integrated into Juniper's products? What's the value of these technologies?

Kriens: This area of application performance is the best-kept secret in networking. [Their products] offer a dramatic improvement in Web site performance and remote access.

Peribit and RedLine weren't well known. Their proposition was to put their respective devices in line and have mission-critical traffic go across them. Their claim was that it was totally transparent. There are now 1,300 users of these technologies, such as ESPN, ABC News, Merrill Lynch. [Now the technologies] will be coupled with our security portfolio and with the routing infrastructure that we have so you move beyond just application performance to application assurance. That's the key.

NW: There's been industry conjecture that you need to make a move into Ethernet LAN switching to truly be competitive with Cisco or Nortel in the enterprise. Is that in the works?

Kriens: No matter what the answer to that question is, one of the crucial requirements is you have to interoperate and complement all of the installed base that's already out there. There are tens of millions of Ethernet ports, supplied by lots of companies. We need to focus on how to bring solutions to that installed base, as opposed to saying the only way to solve a problem is for a customer to throw everything out and only buy our equipment.

It doesn't mean there are not opportunities to enhance the portfolio by whatever means. We do a lot of Ethernet development today within our product line. Ethernet deployment is an increasing percentage of the revenue from our product portfolio. It's a more complex question, but the important thing is to meet user requirements, not just to impose your gear on a forklift.

NW: What else is missing from your enterprise product portfolio?

Kriens: Our main interest is in this strategy of fulfilling our traffic-processing portfolio. Below that are several tactical choices about how to do it. We can partner and joint develop as we announced with Avaya. We can market together with companies; we have dozens of companies with whom we do that. We spend almost $350 million on our own R&D. And we may look at acquisitions. But the important thing is to make the acquisition decision as a subset of the strategy. And people get that turned around. We will always look first and hardest at internal development because that's what has gotten us to where we are. We think we are better at the internal development of the kinds of products we build than other companies of any size attempting the same objectives. Obviously that doesn't mean we don't make acquisitions. But we make them in very complementary ways to fill out strategy.

NW: Most enterprise Ethernet vendors have complementary wireless offerings. Is that something you are looking at?

Kriens: It's just another access technique. Wireless certainly is important to all of us running businesses to access the data and information and applications. . . . We just released the 5GT [wireless firewall], which has lots of wireless security capabilities. Wireless is an increasingly important area and we have a lot of focus on it.

NW: You've put forth the idea that there's underlying dissatisfaction with today's network. Where are you getting this from?

Kriens: I'd call it a technology industry or IT challenge. There are too many people out there that will say whatever it takes to get elected as a supplier. In our discussions with CIOs, when they get together and talk, they are enormously frustrated by empty promises and unrealistic claims . . . such as 'we can do all things, and better and cheaper and faster than everyone else.' People who promise they can do this either don't understand the nature of the business or they think the person they're dealing with doesn't understand. I don't know which is worse.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2005 IDG Communications, Inc.

IT Salary Survey 2021: The results are in