W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1996

Re: Sticky stuff.

From: Jeffrey Mogul <mogul@pa.dec.com>
Date: Tue, 13 Aug 96 13:51:09 MDT
Message-Id: <9608132051.AA06049@acetes.pa.dec.com>
To: Koen Holtman <koen@win.tue.nl>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
    As for the high RTT on some of Digital's internal links: RTTs are
    irrelevant for this discussion.  If anything, they will make the
    savings less noticable.  If Digital's internal links were highly
    _saturated_, that would be another thing.

Precisely.  In the absence of high bit-error rates (such as on wireless
links), high RTTs come from two sources: speed-of-light delays, and
queueing delays.  Although Digital is on a serious cost-cutting kick,
it's not possible to pay less and get slower light, so we use the same
300,000 KM/sec that everyone else has.  But extra bandwidth does cost
money, so in some places we are undersupplied, which means queueing
delays, which means high RTTs.

In other words: high RTTs imply link saturation.

-Jeff
Received on Tuesday, 13 August 1996 14:05:42 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:07 EDT