Want to know why wireless switches are set for sustained, huge growth? Aruba's chief executive, Don LeBeau, was a big cheese at Cisco in the early 90s, and he says the wireless switch market now looks just like the router market did back then. To a hammer, you might say, everything looks like a nail.
"Cisco did $350 million my first year, and $6 billion four years later," he says. His PowerPoint deck includes a slide that shows that Aruba's growth in the last year is the start of a revenue ramp: yes, it's a curve going up, just like the first four quarters of Network Appliance, Netscreen and Foundry (and, we guess, Cisco). And in his world, it seems, curves always go up.
But away from the investor-candy, he has a few things to say about wireless LANs. He believes they are about to become the primary end-user connection in the office, and that wireless LAN companies are set for a central place in IT security provision.
Joining Aruba as CEO in January dropped LeBeau into a whole bunch of technology issues, and he comes up fighting.
The edge will be wireless
The whole edge of the enterprise network will be wireless, says LeBeau, and it will be paid for by old-fashioned IT factors like adds, moves and changes, and reduced port counts . "The economics are overwhelming," he says. Some people talk of wireless as their primary connection method but that isn't enough: "I know a Wall Street bank that says this will be their exclusive connection method."
Shifting to France, he describes a bank with 160,000 employees, and an average of 1.5 to 4 moves per year per employee, with each move costing $50 to $500. That works out at $3000 per employee per year, and a total cost of $500 million, just for adds moves and changes. "Wireless eliminates these costs," he says, because each user is no longer tied to a particular network location.
Whether this is a real bank or an anti-Gallic fantasy, this kind of hand-waving is easy to pick holes in: there are plenty of other solutions that streamline adds, moves and changes (any directory-enabled product pitch majors on it) and while wireless might unshackle a user from a particular port and location, it doesn't do anything to make it easier to change application permissions when his job changes.
But there's more: "Port consolidation in the wiring closet is a huge economic motivation," he says. "That bank has three big wiring closets, with three ports per employee." Put everyone on wireless, and you can support four or five employees for every access point, and get rid of ports. "That bank can go from half a million down to fifty thousand ports," says LeBeau. "The whole enterprise network can consolidate as well."
He brushes aside the objection that in the process users are dropped from the 80 Mbit/s throughput of a switched Fast Ethernet link, to a quarter-share of the 6 Mbit/s throughput offered by a wireless access point: "Where there are high bandwidth requirements, you can still support them with wired ports. Eventually you might need only enough wiring closet ports for access points and servers."
"When there are four or five users per access point, there are no complaints about bandwidth and throughput," he says. And if complaints develop, new wireless protocols will arrive to save the day: "As people feel they need more bandwidth the technology moves on."
It's all very optimistic, and moves him on to Aruba's wireless grid architecture. The wireless grid deploys many access points at desk level, using RF management tools to sort the coverage out, instead of placing them in the ceiling after a careful - and expensive - site survey.
"Wireless was first deployed on a coverage model. Access points were expensive and the goal was to use the fewest access points," explains LeBeau. "In banks it costs $3000 to $5000 to have the access point in the ceilings. You need a union guy to deploy it."
The grid really does eliminate the site survey and RF planning, he says - especially since site surveys are inevitably a snapshot of something that changes over time: "Things alter as soon as those 180lb bags of water that work for you come in the door. RF characteristics will change and change again."
So who came up with the idea of the wireless grid? "It was our customers," he says. The customers said to Aruba: "This is not viable, give us an alternative! Why are we putting the network up in the ceiling when we have structured cabling already? If I'm putting in a wireless network, why the hell am I pulling so many cables?"
Is security your exit strategy?
if anything, Aruba has shifted more towards security in LeBeau's tenure, and the two are tied together: "Security is still the number one concern of IT directors" he says. "Mobility breaks security, and security breaks mobility. We need a solution that embraces security to solve the problem of mobility."
Some of Aruba's customers are more concerned with security than wireless, perhaps: "Thirty percent of our business is with people who want to prevent wireless," he says. That's nothing new. Companies in the wireless space have been selling RF probes as a means to prevent unauthorised wireless access for some time now (Aruba was pitching this when we first met them and before LeBeau came on board). These days, even general networking companies like Foundry, are launching rogue-preventing RF probes.
But LeBeau seems to go further: "We have customers that have no access points or RF monitors at all," [What for?]. Which prompts the question, if there really are customers who see the company as a security vendor rather than a wireless company, is that Aruba's long term goal?
"Security is not a wireless exit strategy," he says, arguing that proper security demands intelligent wireless management. "You cannot address security unless you address wireless and mobility. Even though consumer wireless equipment is a commodity, the wireless grid in the enterprise requires a sophisticated application to manage it."
What about those other vendors?
But why should companies buy from Aruba and not the other wireless vendors? "We're the only vendor with centralised encryption," he says. This is indeed one of Aruba's claims, but it is not clear how important it is.
When Aruba used this as the basis of a security warning in July, it was ridiculed by the rest of the industry. The so-called wireless security problem it guarded against was one which required access to the management VLAN of the wired network.
As well as the technology difference, there's the marketing, of course, and he makes a convincing effort to pour scorn on his rivals' success at getting OEM deals. Trapeze's switches are sold by 3Com and Airespace has deals with Alcatel, Nortel and NEC.
"They must wish they hadn't signed those deals - they have a huge channel conflict," he sighs. "A large OEM distribution is very, very wrong for a high end product. OEMs tend to be big box shifters." The local resellers are the ones with the expertise to do it right he says, and they can't compete with the huge discount the big OEMs get. "Airespace's OEMs are taking the local reseller out of it."
He also has his investors to think of. A VC-funded start-up won't do a successful IPO unless it is doing more than 50 percent of its sales directly or through local resellers. With a lot of big partners, his rivals can't do that, he says.
So, the product is complex and needs lots of support? Well, it's still a high end product, he says, and it is easier to go down-scale in future than to go up-scale.
LeBeau came to Aruba after a period of retirement, since leaving Cisco in 1997. "I thought a father was supposed to provide for his family. I woke up and realised that I had already provided, and I was still going on working 200 percent of a working week." With two children still at school he stopped working and spent time with them, he says - to the extent of coming to Europe for a year, so his son could go to high school in Amsterdam.
He came back into the industry when his children were in college, and looked for a company which combined the two hottest areas: security and wireless. Despite this, he was initially sceptical about Aruba: "Wireless LAN?" he asked. "Why wouldn't Cisco own that?" Nevertheless he realised that Aruba, he says, "is the intersection of security and wireless," and therefore worth his attention
LeBeau admits to having a confrontative style: ""When I arrived, Aruba was the nicest company you could want. But I want fights." he says. "I like friction. You don't get a sharp edge without friction." He wants people to wrestle with ideas, but not to have any religion about things. "I don't have answers: I have my questions,"
LeBeau likes to think his staff are going to take risks, and gives them what he calls a "mistake quota". The idea comes from his Cisco days, he explains: "We needed to push the envelope. The people who were prepared to make mistakes were leading the company. It was never really operationalised, but they had quotas for bookings and revenues - we used the same terminology."
"The key is to make fast mistakes," he says.