DCCP work ideas

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



In response to Pasi's tsvarea talk on where to go next with DCCP, I would like to offer two ideas for future DCCP work that could increase the its likelihood becoming widely used for something.

1. Get involved in the nascent multipath TCP work and develop a way to use DCCP as a better subflow layer, as an alternative to the compatibility "base case" of using legacy TCP connections as subflows.  This way DCCP could become a basic part of a multipath TCP and could get automatically and transparently used to provide the congestion control for standard TCP applications.  For more details see our draft for the upcoming mptcp BOF tomorrow morning (http://tools.ietf.org/html/draft-ford-mptcp-multiaddressed).  Personally I see this by far the most promising potential use for DCCP, but I admit some bias. :)

2. In the interest of addressing the comment after Pasi's talk about both DCCP and its potential applications wanting to have "control" over congestion control...  Part of the problem seems to be that typical multimedia applications operate at only certain "quantified" bitrates: for example if an application tries 128Kbps but DCCP throttles it due to congestion, it has to cut all the way down to its next lower speed of, say, 64Kbps and stay there, even if say 100Kbps is actually available.  The problem is if the application keeps just sending 64Kbps, DCCP's congestion control will not try to probe for bandwidth and the application will never know when it can move back up to 128Kbps.  So solve this by developing an extension to DCCP's congestion control mechanisms an a corresopnding API allowing applications to maintain a standing "request" for more bandwidth than they're actually using at the moment, and to notify the application when the full amount of requested bandwidth appears to be available.  That should allow media applications to follow DCCP's congestion control decisions without giving up the control they need in order to utilize available bandwidth dynamically.  There are several alternative ways to achieve this at the congestion control level, at least one of which might even be reasonably safe and efficient; I'll try to write it up in a follow-on E-mail shortly.

Thanks,
Bryan



[Index of Archives]     [Linux Kernel Development]     [Linux DCCP]     [IETF Annouce]     [Linux Networking]     [Git]     [Security]     [Linux Assembly]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [DDR & Rambus]

  Powered by Linux