Open letter to Internet Engineering Task Force: Back off Cisco, not all members want to 'play to your tune'

A mini revolution against Switchzilla takes form in a corner of industry group

6 Reg comments Got Tips?

There is growing unrest at the Internet Engineering Task Force (IETF) due to the power that some accuse Cisco of wielding over the global community of network designers, operators, vendors and academics.

The open standards IP association was set up to oversee the “evolution" of internet architecture and the "smooth" running of the net. It includes steering groups, directorates, affiliate groups, trusts and so on.

Members include big businesses from the world of networking - Juniper, Ericsson, Nokia, F5 and, of course, Switchzilla, as well as companies involved in ops management, security, applications and research.

There is, however, frustration from some IETF participants and they have made their professional opinions known - in a personal capacity - this week in open letters posted on LinkedIn, co-authored by Andrew Alston, group head of IP strategy at Liquid Telecommunications and a Juniper ambassador, Sander Steffan, IPv6 strategy and design consultant, Fernando Gont, security researcher at S16 Networks and Jan Zorz, CEO at Go6 Institute.

The four didn't name this omnipotent force throughout their entire post on 15 April, but it seems clear which firm they were referring to when they said it is the most dominant player in routing: Cisco has market share of more than 50 per cent. And they claimed Cisco is ruling the roost at IETF too.

"This has led to forcing every other IETF contributor to play their tune whenever that vendor has a stake, or risk having their proposals being left to wither on the vine," they added. Examples of this have become "endemic" in the Source Packet Routing in Networking (SPRING) Working Group and "we see similar happening elsewhere is IETF."

It is only a matter of time "before we see this behaviour creeping into IDR (Inter-Domain Routing), LSR (Link State Routing) and other working groups. It seems that the will of operators/ other vendors is being left behind," the letter added.

The four asked the big networking brands if they intend to continue to let this happen.

"We are looking at you Juniper, Huawei, Nokia, Ericsson, ZTE, and so the list continues - are you content to continue to have your future, and the future of your customers dictated by one vendor in the routing space?

"Are you prepared to let your tomorrow be dictated by the whims of a single vendor? Where thoughts and ideas are all entirely dictated by one vendor in the routing space? Are you prepared to have your own innovation hampered or otherwise ignored unless you can get the buy in of ONE vendor? Are you ready to admit it?"

Should members of IETF say "enough is enough and demand a standards revolution?" they asked, before ending the post by saying the Internet was "built from the bottom up" and this is "precisely the approach we need to return to, instead of the interests of a single vendor based on their stock performance."

The IETF board is comprised of 13 people and the chair is Alissa Cooper, who is also a fellow within Cisco's Collaboration Technology Group. Others include Martin Thompson, distinguished engineer at Mozilla; Gabriel Montenegro, principal architect/ developer/ program manager in networking tech at Microsoft; Ted Hardie, who describes himself as an Internet Boffin at Google; and Jari Arkko, senior expert at Ericsson Research.

Most of Cisco's routing rivals don't seem to be represented at the top table, which may be partly why some feel their voice isn't heard to the same extent.

Steffan told The Register that "it is getting impossible to get anything done" in the areas he is involved with IETF "without support from Cisco".

"There are a LOT of Cisco employees in important positions in the IETF, and they are getting more arrogant and blunt. A chair who won't call for adoption of a draft because they don't want it, no matter what the working group would think."

He said in SPRING, for example, Cisco has declared consensus of documents that contain "LOTS of open questions and concerns. You either rubber-stamp what Cisco says or you get ignored," he claimed.

Next steps

In a separate post, published on LinkedIn yesterday, the same four people said IETF has "performed an extremely valuable function" over the years in working toward the common good of the Internet and "many sections function well".

However, one of the fundamental flaws is the way corporate governance has evolved, they claimed, and to deal with what they see as "conflicts of interests," came up with some suggestions. These include shifting the structure so the IETF working group chair and area directors don't work for the same employer. "It creates potential conflict - or at the very least the perception of potential conflict and bias - real or imagined."

Further, the appeals process is "flawed" and needs to be handled by a "third body entirely," removing it from area directors and working chairs; the recall of working group chairs should not be at the "sole discretion" of area directors; and there should be rotation of working group chairs so that no chair is in situ for longer than two years, with two chairs per working group. "[T]here should never be a situation where a working group chair can act unilaterally," they wrote.

The Reg asked Alston to comment further but he declined and Cisco also said it had no comment at this stage. We also asked ZTE, Huawei, Arkko at Ericsson, Nokia and Juniper to comment. We have yet to hear from those companies.

This isn't the first time IETF has faced criticism, far from it. Renowned Internet Hall of Famer Randy Bush, wrote back in 2005 that "many researchers" saw IETF as an irrelevance because they were either "measuring the real internet as a behavioural phenomenon" or "wanting to do research 'beyond' the internet. Neither involves the IETF".

Bush's two-page report, Into the Future with the Internet Vendor Task Force: a very curmudgeonly view [PDF], said vendors blocked technical advances that were needed in standards.

"The IETF," he wrote, "has grown so large and so enamored of complexity and featur-itis that it is a full-time job to participate. Who can afford to spend full-time on IETF? Vendors with more features to add to sell more baroque systems. It's not an evil conspiracy, but rather a consequence of being enamored of complexity and the vendors' need to keep selling 'new', 'better' products."

He added: "The operators are the wall. And they pay capital costs and operational expense to deploy complex features which vendors market as needs to the users. And then everyone wonders why the margins went down and the prices stayed up".

He said a "social vision of openness, fairness and inclusion" had taken over from "rigorous design and development" of protocols.

"One has only to look at the long and painful histories of (not) deploying IPv6, dnssec, various QOS schemes... to see the results of the disconnects between research, the IETF, process and the realities of operational deployment. "

The "real engineering vision" had "fallen out" of the IETF, he concluded.

Some research that Zorz at the Go6 Institute conducted in 2014 looked into why operators, particularly small to medium-sized ones, were not joining IETF mailing lists or attending meetings, and so leaving vendors, academics and large businesses "to rule many decision making processes within the IETF."

The answers included time, money, culture and awareness. It seems that all these years later, some systemic challenges remain within IETF. ®

SUBSCRIBE TO OUR WEEKLY TECH NEWSLETTER


Biting the hand that feeds IT © 1998–2020