1/22 The problems with the Cory/Soda networks seems to be pretty
widespread throughout campus. Is the routing being handled by
monkeys or something? It's soooooooooooooo slooooooooooooooooooow.
\_ Since about a week-and-a-half ago, the campus has been hitting
its bandwidth cap on its link to the commodity Internet
(70 megabits/sec). About half of the traffic is kazaa/gnutella.
CNS and the chancellors are trying to decide what to do about it.
-tom
\_ So when did this bandwidth cap start, and why didn't CNS
announce it in http://ucb.net.announce like they do all other service
interruptions? This reminds me of when then initiated the
cap on the dorms and wouldn't tell Residential Computing what
the "problem" was for 2 days. -mikeh
\_ The bandwidth cap has been in place since our OC12 was
installed--it's only recently been hit. 70 megabits/sec
is all the campus is paying for. "bandwidth cap" is
somewhat misleading; we're talking about the size of
our pipe. It wouldn't be any different than if we had
a 70 megabit/sec pipe. Our current hardware connection,
if we let traffic run up to it, would give us about another
15 megabits, at a cost of $4500/month. -tom
\_ That's easy. Cut off the fucking dorms from anything but a web
proxy/cache and locally hosted email.
\_ "Every problem has a solution that's simple, elegant, and
won't work." The dorms are already limited to a separate
40-megabit cap. -tom
\_ So total campus bandwidth is 110? Ok, ban it and start
executing p2p users. It's all theft anyway.
\_ Bandwidth to the commodity net. Internet 2 traffic
is not capped, or tapped out at this point. (I think
our pipe is something like 650 megabits). -tom
\_ just cut off their mouse hand.
\_ And in other news, masturbation drops 80% on
campus, survey says.
\_ 70mb/sec =~ 8.7MB/sec?
\_ So the 400ms+ ping roundtrip times are the result of a
bandwidth cap?
\_ they are the result of us hitting the limit of the bandwidth
we are paying for, yes. -tom
\_ I suggest taking this to http://ucb.net.discussion |