IETF 75 and the ISOC Fellowship
The ISOC Fellowship
In Feb 2009, I won an ISOC fellowship to attend the IETF conference. The ISOC fellowship pays for the airfare, stay in the hotel and the conference fees of the IETF participants. The process to apply for the ISOC fellowship is competitive and only participants from third world countries can apply. In Jan 2009 there were 145 applicants who wrote proposals for the work they plan to do at the IETF out of which 4 people were selected. My proposal was regarding Internet Measurement and Analysis – the area of work I have been working at Akamai for the last 5 years. More specifically I was working on drafts which outlined the metrics and the ways for measuring them for websites/webapps and streaming quality. Network application measurements has been one area which has been weakly represented and not much work was happening in these area. There was some related work happening in IP Performance metrics groups and PMOL working groups but none in the above mentioned areas. My work was to fill in those gaps by writing drafts and then soliciting comments and finally working through the IETF process to get it standardized. I was also vouched for by two people who knew about my work which helped me win this fellowship.
The IETF and conference details
The IETF conference is the topmost conference where network engineers meet to discuss RFCs, best practices and share operational knowledge of running various that make up the Internet. I attended the 75th edition which was held in Stockholm from July 26 – July 31st 2009. The IETF is different from most other conferences in the sense that it is a collection of working group meetings. Each of the working groups decides it’s own agenda and the drafts that they want to discuss in the meeting but bulk of the work happens on the mailing list beforehand. Also anyone can join the mailing list and participate in the discussions. So the IETF meeting really is a loose collection of several miniconferences (for the lack of the better word). Please read the Tao of the IETF if you are interested in how the IETF works. The following quote (from the Tao of the IETF) quite accurately describes how the IETF works.
“In many ways, the IETF runs on the beliefs of its participants. One of the “founding beliefs” is embodied in an early quote about the IETF from David Clark: “We reject kings, presidents and voting. We believe in rough consensus and running code”. Another early quote that has become a commonly-held belief in the IETF comes from Jon Postel: “Be conservative in what you send and liberal in what you accept”.
In that sense there are strong parallels (low barrier to entry, open access, easy to contribute, dominated by hands-on engineers) with the Free/OpenSource world as Linus Torvalds said
“Talk is cheap. Show me the code.”
If you participate in the IETF, be prepared to have your draft (and data as well) questioned, scrutinized, critiqued and commented upon. It requires lots of patience (and a thick skin 
, convincing power with proper data to prove what you have proposed is worth standardizing and being accepted as an RFC.
My Experience
The IETF conference attracts the best and brightest in the field of networking. It was a pleasure to meet so many great engineers. It was easy to talk to most of other engineers as they were very down to earth. There was no ivory-tower syndrome.
On Sunday, I attended the first-timers meeting. It was useful as it gave an overview of what to expect at the IETF meeting and how to make the best use of it. On Monday, I attended the Audio-Video technology working group meeting. An engineer from Huawei was presenting a draft and participants asked him several questions but he could not answer. When asked about the use-cases when scaling to hundreds of thousands of users, he said that they had tested for only 3-5 users. On hearing this one of the participants said sarcastically that we design protocols that scale for millions of users and not just a few users. This was first of many such blunt and sarcastic comments that I had heard during the conference. It was clear from then on that, if you have not carefully considered all the cases, the draft would not go for further review and will be shot down mercilessly. The next meeting was the Benchmarking Methodology WG. IT was a small crowd as compared to the AVT WG and the proceedings were more cordial. There was some remote participation in this meeting. Personally I found this meeting useful as it was related to my area of network application metrics.
On Wednesday, There was a PMOL (Perfomance Metrics for Other Layers) BoF Session. I talked about perfomance metrics for Webapps and streaming at the PMOL BoF and got some good comments from Al Morton (my mentor) and Henk Uijterwaal (WG Chair for IPPM WG). Al later helped me iwth the xml2rfc tools that are needed to format the RFC draft. Overall the response was positive and I got some good pointers to some earlier work in other standardization bodies such as the ITU. On the subsequent days, I attended the Operations & Administration and Technical plenaries and a talk on “Securing the DNS”.
On Friday, I attended the IPPM (IP Performance Metrics) working group meeting. I was really well-prepared this time having attended several WG group meetings. Also since I had read the drafts and the background materials on the drafts that were going to be presented, I was able to make meaningful comments on some of the drafts that were being presented.
Among the other highlights of the meeting were having a photosession and one-on-one chat with the Vint Cerf – one of the inventors of TCP/IP and the Internet and the welcome reception at the City hall in Stockholm where they have the Nobel Prize banquet. It was addressed by the mayor of Stockholm among others and it felt great to occupy the same space as some of the brightest minds of this century (of course, separated by time .