On 8 Apr 2011, at 12:13, Adrian Kennard wrote: >> ... >> Does anybody have any stats on the kind of packet scheduling >> algorithms employed between/at DSLAMs and backhaul networks? >> >> Background: >> >> I'm trying to weigh up the arguments for/against “identified >> application type”-based “traffic management” systems on consumer >> broadband connections, versus something altogether simpler and >> straightforward and—at first glance—altogether more “fair”, such as… >> fair queueing. >> >> Expert opinions would be much appreciated, both on and off-list. > > Biggest issue is the likes of bit-torrents. They are lots of separate > tcp sessions so normal "full link" behaviour tends to balance between > tcp sessions not between customers. Something more aware of per-customer > traffic levels and management is likely to be fairer, or something > specific to the types of traffic like this that are "unfair" somehow. >
On the grounds that it's not possible for an operator to know how their subscribers value the various applications they're using, protocol or application 'agnosticism' is critical. http://tools.ietf.org/html/rfc6057 describes one way of delivering this that is in use today. Mat
