Author Archive: Guido Appenzeller

Best Paper Award for Neda, Yashar, McKeown et. al

Congratulations to Neda, Yashar, Monia, Nick and Geoff for their best paper award at the Internet Measurement Conference. Their paper Experimental Study of Router Buffer Sizing tests out recent results on  buffer requirements of high-speed routers that serve highly aggregated traffic. Amongst other things it verifies the C/sqrt(n) result from my thesis as well as my former office mate Yashar Ganjali’s work on very small buffers and find that they hold well.

It is great to see this work getting recognized, but what is even more encouraging is that two router vendors privately confirmed to me that the next generation of some of their products will have substantially smaller buffers. This not only reduces power consumption, but also means that we are less likely to see latency spike whenever peering points or core links are congested.


The project that currently takes up the majority of my time at Stanford is OpenFlow. OpenFlow is a new protocol that we specified and that vendors are now adding to their routers and switches. What OpenFlow allows you to do is remotely control the behavior of a switch from an controller software that runs on a standard server. This has two major advantages:

  1. You can now write your own control software and try out new switch functionality at full line rates. In the past this has been difficult as all major router and switch vendors lack APIs and are typically closed platforms.
  2. If you use a centralized controller now has a unified view of the network. For some applications such as mobility management, virtualized data canters or security this allows you to do things that previously would have been very difficult or impossible.


Pacific Coast Trail Runs

This week-end I did my first run that was organized by Pacific Coast Trail Runs, and it will definitely not be my last. They label themselves “Serious Fun”, which is pretty much accurate. The runs are less focused on racing, and more on enjoying the run itself. However the distances of their runs are “serious”. Some events have as the minimum distance marathon or 50k, and some entrants run pretty crazy distances.

I ran the Santa Cruz Mountain run and it was a blast. The 21k course involved wading through the 3 foot deep San Lorenzo River, climbing over fallen logs, waiting at the tracks of the Santa Cruz Railway for it to pass and 2,300 feet elevation gain. I also love their aid stations. Water melon, pumkin pie and Chili at the finish line. If you like running medium and long distances, you should definitely try them out.

SIGCOMM Best Demo Award and other News

Currently the main thing I am working on at Stanford is the OpenFlow Standard. OpenFlow basically allows a software that is running on a commodity PC, to remotely control the flow table and thus the entire routing functionality of a switch (see the excellent white paper). This is a very interesting application for networking, as it allows researchers or start-ups to build new technologies in this space. Think of it as the Facebook API for the networking world. If it takes off, it will lead to a lot of innovation in this space.

Last week we presented a demo of using OpenFlow to migrate game servers in real time (and while people play games on them) across different IP subnets at SIGCOMM 2008. The demo received rave reviews and won the Best Demo Award by a wide margin. I will post more about the Demo and other OpenFlow news on the OpenFlow Blog soon. This week we also held the first OpenFlow tutorial at HOT Interconnects at Stanford (organized by our own John Lockwood this year – great work John!). It was well received and we learned some good lessons how we can improve it in the future.

Congratulations Sundar and Seth!

The 2008 MIT Technology Review list of top innovators under 35 came out today,  Sundar Iyer and Seth Hallem, both fellow Stanford Ph.D. students, made the list this year, and I am very happy for both of them. Sundar has had a profound impact on how buffers for high-speed routers work, and Seth has changed what is considered best practice for code checking. Both have also founded great start-up companies in the process.

Congratulations Sundar and Seth!

Update: As Sundar points out, Stanford CS faculty Andrew Ng and former Ph.D. student Meridith Ringel Morris also made the list. Congratulations here as well!

EIR at Morgenthaler Ventures

A few people mentioned that they were suprised about my move to Stanford after leaving Voltage. This post should be a lot less suprising: I am joining Morgenthaler Ventures as a part-time Entrepreneur in Residence. Spending a day a week at Morgenthaler Ventures is a very synergistic combination. The exposure to bleeding edge technology in Stanford helps understanding entrepreneurial opportunities. Vice versa seeing the hard technical problems that portfolio companies face helps define meaningful new research areas. And while some of you may not have heard of them before, Morgenthaler is a great firm to work with.


Joining Stanford

Hector Garcia-Monial and myself at Graduation.I am happy to announce that this week I have joined Stanford University as a Consulting Assistant Professor. This may come as a suprise to some people, as I am not exactly your typical academic. Those people would be correct, my job here is not primarily about teaching. The main reason I am joining Stanford is OpenFlow, and it is one of the most exciting technologies I have seen in networking for a long time.

OpenFlow is exciting in two ways. First, it allows you to run new protocols and algorithms on production networks. Before OpenFlow this was very hard, as modern routers have no API that gives access to this low level functionality. Second, it allows you to make centralized yet fine grain routing decisions. This has huge advantages in some areas such as security, data centers or mobility.


Moving on…

Friday last week was officially my last day as an employee at Voltage. The company has been doing great, both in terms of the organization as well as market traction. However after working there for 6 years it felt like it was the right time to move on. Since we founded it in 2002, Voltage has developed from being an early stage technology venture, to a stable, self-sufficient leader in enterprise software. And this means I can move on to do something new.

Early Voltage Team

Building a company is primarily about finding the right people and building the organization. Conversely, the in my count most common cause of death for an early stage start-up is dysfunctional team. At Voltage, we started with a founding team that shared a common vision and terrific investors that were an incredible help whenever we needed them (picture of the early days on the right). On these foundations we together built the great organization that Voltage is today. I am grateful for having had the opportunity to work with everyone in the company, from the executive team, to the board to each and every employee. In retrospect there are only few hires that I would not make again. Going forward I am confident we have the team in place that has what it takes to guide Voltage into the future.


Server Back up Again

As you might have notices, this blog as well as all email was down for the past few days. The reason was an explosion in The Planet‘s Houston H1 data center, that took out power for our server and about 9,000 other servers as well. More information here. About 700,000 web sites were reported off-line, amongst them b3ta. Assuming the back-up generators hold up, things should be stable for now.

All of this seems to be another good argument for virtualization. While I am not sure how well something like EC2 is set up for an equipment failure like this, at least in theory having multiple data canters and a distributed S3 would allow you to migrate the virtual machines to another center.

Traveling Light with Children

Traveling with Baby and Hand Luggage OnlyAs any frequent flyer knows, there are two types of luggage in air travel: cabin luggage and lost luggage. Sometimes lost luggage is found, and delivered to you eventually. Sometimes all you get is an apologetic letter and a check in the mail three months later. That happened to me last year with British Airways, but then again I was asking for it by flying through Heathrow Airport.

Today I almost always travel without checking bags. I traveled to my wedding and honeymoon only with hand luggage, and managed to get our Kilimanjaro trekking gear into the cabin as well.

Recently someone claimed this would have to change once we go traveling with our first kid. Well, so far I still think hand luggage is the way to go. As you can see in the photo from our two and a half week trip to Europe, we didn’t even have to resort to oversized “personal items”. Just two bags (which also held our two laptops) plus a small one for Daniel were sufficient. We did order formula and diapers in advance, and did laundry locally (for the US, check out Babies Travel Lite). The hardest part was to organize an infant car seats for the taxi to the airport and the rental car in Europe, but even for that services are emerging.