FEATURE SERIES: Network Trailblazers
Network Trailblazers: A Conversation with Kirk LougheedElizabeth Corcoran View All Contributing Writers 06/13/2011
Network Trailblazers is a new series that highlights the creators and visionaries of the Internet network.
The underpinnings of Cisco's products--and its corporate culture--were laid in 1984, in large part by founders Len Bosack and Sandy Lerner, and Kirk Lougheed, the company's first engineer, who today is a Cisco Fellow. As part of Stanford University's computer support team, they started with a simple premise: serve the customer.
Like many organizations, Stanford's computing resources looked like an alphabet soup of different systems--independent Ethernet islands. Those systems needed ways to communicate directly, rather than relying on the Arpanet. Several researchers at Stanford began building router technology to exchange PARC Universal Packets (PUP).
Lougheed and his colleagues realized that Internet Protocols were fast on their way to becoming the common language of data exchange and rewrote the technology to rely on IP. Finally, they appreciated the power of simplicity. Although the software that they developed, which became Cisco's Internetworking Operating System, or IOS, lacked some of the niceties of more mature software, it could be swiftly adapted to serve customers' needs.
Lougheed talked recently with journalist Elizabeth Corcoran from his office at Cisco's San Jose headquarters and described how Cisco's core principles--serve the customer, rely on IP, and adapt to new demands--emerged and how the world has changed as a result. Here are highlights from that conversation.
"The basic protocols are the same ones as back in 1984. It's figuring out how people use the network--that's where we've really changed."
Who were your first customers?
One of our first commercial customers was actually HP Labs. They regularly sent a courier down between Palo Alto and Cupertino with a mag tape, delivering data back and forth. So when we delivered two routers with a 65-kilobit link between them they thought that was absolutely fantastic. Basically we were eliminating the "sneakernet" style of communicating between machines.
What drove the early development of IOS?
After we had left Stanford, we started asking what do the customers want? They wanted to build large scalable networks. We introduced our routing protocol. Customers had computers that spoke not IP but other protocols. We worked the software so it would support these multiple languages as well as IP. So we could hook up a DECnet computer, an Apple, Xerox protocols, as well as IP and run them across the network.
Did your early sales team ever over promise what you felt you could deliver?
In my particular instance, a VP of sales, went off and visited the financial district of New York. He came back and said, "This network has to reconfigure itself in, oh, say, on the order of seconds." And I just looked at him and said, "It doesn't work that way." And then the next morning, I came back and said, "I think I know how to make it work that way." Engineers will typically respond that way the first time--then the brighter ones go off and think about it.
What are the greatest strengths of the network--and its greatest vulnerabilities?
The network's greatest strength is that it ties people together. It's weakness? We're very dependent on the network. If you took the network away from people for a week or so, they'd have real severe problems, not to mention the economic problems. When Egypt was having its revolution and the Egyptian government unplugged Egypt from the network, they were predicting all sorts of chaos if they weren't plugged in by Monday morning when the banks started opening. So we are very dependent on having the networks work.
That's our weakness--not the network's. What about the technology itself--is it vulnerable?
It is, but the network has been under almost 24-hour-attack for past 20 years. Almost every time somebody tries to poke a hole in it, we figure out how to correct the flaw, and by "we," I meant the industry. So I'm not terribly worried about structural flaws in the network; it's under stress all the time and we know how to keep it up and keep it running.
With the benefit of hindsight, what would you have done it differently?
I--and every other engineer--have a long wish list of, "If only he had done things differently!" On the other hand, IOS is a very simple piece of software for engineers to extend. Because we kept it simple, we were able to do all sorts of things quickly and respond quickly. If we put in a lot of the safety stuff or used better engineering practices early on, it might actually have slowed us down, slowed the growth.
So you mean the network is a good example of what Clayton Christiansen describes, namely that big innovations don't start out perfect?
They aren't perfect. There's always something missing.
What will future customers want?
They'll want more bandwidth; they'll want it everywhere. They'll want the network to be always up. They might start to expect it in all sorts of oddball places, like, in the middle of the Nevada desert. And they'll expect all their stuff, all their digital stuff, to be on the network. And with more bandwidth and processing speed, there might be interesting things you can do with virtual reality; you will be able to adapt as fast as the human mind can adapt.
What about the network has surprised you the most?
That it's pervaded every aspect of life. That it's always there. There is an entire generation of people, like my daughters, who have only known life with the network. They can't imagine life without it.
What's changed more: the technology of the network, or the world because of the network?
The world's changed more because of the network. A lot of the stuff in the network was figured out fairly early. The basic protocols are the same ones as back in 1984. It's figuring out how people use the network--that's where we've really changed.
One tool you would have wanted if you were marooned on a tropical island in 1986:
Assuming that I had food and shelter, I'd want a set of books.
One tool you would want if you were marooned on an island today:
A computer with an Internet connection.
Favorite example of a prediction you heard early in life that hasn't come true?
I haven't spent my life waiting for my science fiction stuff to come true. But one thing that disappoints me is that we don't have more of a space program. The moon was great when I was 11 years old. I would have expected that we would have gone much, much further in the intervening years.
Favorite prediction that turned out to be true beyond your wildest imagination:
How useful this Internet stuff turned out to be.
Advice for someone heading to college today:
I told my daughters to study what they are interested in. One's a political science major and one's a classics major. Follow your dream, absolutely.