[e2e] Is a non-TCP solution dead?

[Nicolas Christin]

On Fri, 25 Apr 2003, Saad Biaz wrote:

> Currently, numbers says about 90% of the traffic is TCP. If TCP is
> so bad and slows down too badly for no good reason, then the link
> utilization of backbone routers must be quite low.

Saad,

I agree that numbers would be needed, but...

Actually, backbone utilization is indeed low - see a number of papers
coming from the IPmon project at Sprint labs. However, this is not
because of poor TCP performance, but instead, because backbone links are
generally overprovisioned. Well, that's at least the case on the Sprint
network, which is the only place I know to have published *recent*
numbers. Any other reference is welcome.

> Please, some numbers!!!

I am afraid numbers of backbone link utilization will not lead any
insight w.r.t. TCP performance, because of overprovisioning. On the
other hand, I'd be *extremely* interested in seeing *proof* that non-TCP
(or "fake" TCP) traffic is increasingly dominant, as Alex was initially
implying, before seemingly backing off in his last post.

Based on discussions with a number of people, and judging from the
number of universities which are trying to ban peer-to-peer traffic
because it is clogging their access links, I believe that the major
shift in traffic patterns observed in the last couple of years boils
down to the emergence of peer-to-peer traffic, far more important in
proportion than the deployment of multimedia applications, VoIP, etc.
Well, quite a number of peer-to-peer applications (e.g., Kazaa, WinMX)
seem to tunnel their traffic through HTTP (I am sure about WinMX, unsure
about Kazaa) precisely to be able to go through proxies, firewalls, etc.
These data transfers appear to follow regular TCP congestion control -
to the extent the host stack is complying. They don't appear to just be
"posing as" TCP, they are TCP flows.

If anyone can back this with actual numbers, to confirm or deny what
I just wrote, that would be appreciated. But just stating that there
is a drastically increasing number of non-TCP connections is quite
counter-intuitive (per the argument above) and requires proof.

The other problem I see with this thread is that the argument seems to
change every time a new post is made. From what I read, "TCP performance
sucks" became "Traffic is becoming non-TCP/fake TCP" before rolling back
to "TCP performance sucks". OK, fair enough, Alex, but please back this
argument with *data*, otherwise, it just reads like a disorganized rant.

Regards,
-- 
Nicolas