North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

ECN

  • From: Mikael Abrahamsson
  • Date: Fri Nov 07 02:28:09 2008


Hi,


On LKML (Linux Kernel Mailing List) there is talk <http://lkml.org/lkml/2008/11/4/151> about shipping the Linux kernel with ECN turned on by default (it was on by default a few years back but that change was reverted due to too many sites dropping ECN enabled SYNs).

Recent investigations <http://www.imperialviolet.org/binary/ecntest.pdf> shows that 0.5% of end hosts will drop SYN packets with ECN turned on. This is approximately the same rate I have seen for A/AAAA adoption in this tread <http://www.ops.ietf.org/lists/v6ops/v6ops.2008/msg01585.html>.

Do we in the operational ISP community have an opinion on ECN adoption that we want to voice? As far as I can discern from <http://www.cisco.com/en/US/docs/ios/12_2t/12_2t8/feature/guide/ftwrdecn.html> for ECN to actually be useful, we (the ISPs) have to turn this option on in the routers as well. Is anyone doing this today? What vendors support it?

When I thought about it, the IP core (10G links etc) first came to mind, and there it's fairly easy to roll out (since I guess a lot of us do WRED already), but what about on slower links? Would it make sense to have our DSLAMs do this? What about DSL/cable modems (well, vendors should first realise that FIFO is not great to begin with :P) ?

<http://www.icir.org/tbit/> is a summary page I found on ECN that looks like a good resource for further reading. Is anyone looking into including ECN configuration into some BCP document?

--
Mikael Abrahamsson    email: swmike@xxxxxxxxx