Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Bufferbloat is caused by buffering hundreds of packets, not 10. Bandwidth-delay product has increased so much that the initial congestion window also needs to increase so that TCP can ramp up in a reasonable time.


Their own paper shows that it's still a controversial issue and after a certain point decreases performance.


This paper seems unreservedly in favor of larger windows. http://research.google.com/pubs/pub36640.html

Based on our large scale experiments, we are pursuing efforts in the IETF to standardize TCP’s initial congestion window to at least ten segments. Preliminary experiments with even higher initial windows show indications of benefiting latency further while keeping any costs to a modest level. Future work should focus on eliminating the initial congestion window as a manifest constant to scale to even large network speeds and Web page sizes.


I could be reading it wrong, but I think I see some issues:

https://docs.google.com/gview?url=http://www.cs.helsinki.fi/...

   IW10, while improving elapsed times, imposes higher queuing delay than IW3
   However, if self-congesting, IW3 is more aggressive in terms of queuing delay
   AQM (RED) failed to control the increase in the queuing delay




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: