IAB Workshop on Interdomain Routing (proposal)
All protocols accrue new functionality over time as a result of new use cases, new operational requirements, and new ideas about how to address existing problems. BGP is no exception. At some point, however, a single protocol can accrue enough functionality, across a wide array of use cases, to reach an undesirable level of complexity for effective operation. Some hold BGP has reached a state outside RFC1925 rule 5, becoming more complex than necessary to meet its original objectives.
For further information on the perception of BGP's expansion, the complexity of the protocol, and security, see [problem statement].
Further, pressure is rising to provide a "more secure" interdomain routing solution, whether by adding new functionality to BGP, by augmenting BGP, or replacing BGP as the dominant interdomain routing protocol.
Towards these ends, the IAB is holding a technical workshop to bring researchers, operators, and protocol designers into a conversation about how to classify, understand, and resolve current problems with interdomain routing.
Goal
The goal of this workshop is to facilitate a community discussion, bringing a diverse array of resources to bear on classifying, understanding, and resolving the problems outlined in [problem statement], potentially including:
- determining how to incementally modify BGP to support all use cases in simpler ways
- determining two different protocols should be used for IDR and services, and understanding how to divide these use cases among multiple protocols
It is always tempting to simply "replace BGP," but this (probably) isn't a realistic goal. Instead, this workshop is intended to begin a conversation about how to think about the problems facing BGP, and propose research directions for resolving these issues.
Format
This workshop will be held on [xx] days in a virtual format. Proposals will be accepted until [xx], after which time the program committee will determine:
- Papers accepted for presentation
- Papers accepted for publication on the workshop website
- (Discussion) Panels accepted for presentation
Workshop attendance will be by invitation only, based on papers or panels proposed, paper or panel acceptance, and community activities.
Proposal Guidelines
Proposes should directly address one of the following:
- Support or contradict [problem statement], giving specific research or evidence
- Discuss a solution for one or more of the issues raised in [problem statement]
If a proposal includes a solution to a given problem, they should include the transition mechanism from the current state to the proposed state and the economic drivers which would move the community from the current state to the proposed state. Solutions, in other words, can be research or ideas in formation, but must not be purely technical; they must take operational and economic drivers into perspective. Research proposals should be concrete, giving the specific data to be used, the way in which the data will be collected, how the data will be handled and processed, and the desired result of the study.
In short, this workshop is eliciting conversations around practical solutions--even if they are not fully designed or developed--with a path to deployment, as well as opening a discussion on [problem statement].