1 / 8

Active Queue Management (AQM) in the IETF

Active Queue Management (AQM) in the IETF. IETF 86 - TSVAREA Orlando, Florida March 13, 2013. Goals. In late 2012, the TSV ADs asked the TSVAREA list ( tsv-area@ietf.org ) for some feedback on possible AQM work in the IETF We got some feedback (THANKS!)

aleda
Download Presentation

Active Queue Management (AQM) in the IETF

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Active Queue Management (AQM) in the IETF IETF 86 - TSVAREA Orlando, Florida March 13, 2013

  2. Goals • In late 2012, the TSV ADs asked the TSVAREA list (tsv-area@ietf.org) for some feedback on possible AQM work in the IETF • We got some feedback (THANKS!) • This presentation summarizes what we think we heard, and asks for more discussion

  3. Background • Buffers exist, some possibly large • AQMs can improve performance for traffic hitting a bottleneck, especially real-time interactive traffic competing with loss-based traffic • Recent AQM proposals: • CoDel: draft-nichols-tsvwg-codel • PIE: draft-pan-tsvwg-pie • Recent TSV AQM work: • draft-ietf-tsvwg-byte-pkt-congest • But AQMs are not exactly protocols, and general AQM work may not be clearly in-scope for existing WGs

  4. Should we have a WG? • This depends on: • If there’s energy to do “something” • Where “something” means either BCPs or Proposed Standards • Multiple existing paths exist to do Informational/Experimental documents if there isn’t this level of confidence and support for ideas (ICCRG, AD-Sponsored, Independent Stream) • Assume if there’s anything worth doing, that it’s too much to bolt onto TSVWG’s charter

  5. Should we have a WG? • This depends on: • If there’s energy to do “something” • Where “something” means either BCPs or Proposed Standards • Multiple existing paths exist to do Informational/Experimental documents if there isn’t this level of confidence and support for ideas (ICCRG, AD-Sponsored, Independent Stream) • Assume if there’s anything worth doing, that it’s too much to bolt onto TSVWG’s charter • We heard: • Possibly a WG would be productive; some people at least are interested and think this is timely. • At least one person though it belongs in INT or OPS

  6. What would the WG do? • Algorithm specs as BCPs / Standards Track? • General requirements and design space analysis for AQM algorithms? • E.g. behaving well under load, being efficient to implement, etc. • BCPs for configuring “legacy AQM” like RED?

  7. What would the WG do? • Algorithm specs as BCPs / Standards Track? • General requirements and design space analysis for AQM algorithms? • E.g. behaving well under load, being efficient to implement, etc. • BCPs for configuring “legacy AQM” like RED? • We heard: • Lots of thoughts … some see value in specs, at least • Setting the bar for published specs may require some other work too, e.g. on requirements, how to evaluate algorithms, etc.

  8. Summary & AD Thoughts • Should we have a WG? • Probably yes, if there’s rough consensus on 2nd question! • What would the WG do? • Aim for one or more specifications • If there’s energy, or lots of proposed algorithms, do requirements, test methods, etc. • Need more feedback, and to identify proponents that are willing to do the work, chair a WG, possibly hold a BoF, work with other areas (e.g. OPS, RAI), etc.

More Related