Article View: crl.general
Article #50Re: Interconnect speed
From: Matt Dillon
Date: Thu, 08 Dec 1994 02:38
Date: Thu, 08 Dec 1994 02:38
48 lines
3181 bytes
3181 bytes
:In article <3c5m64$584@crl10.crl.com>, Eric Penn <stupid@crl.com> wrote:
:>
:>
:>I've seen some people posting output from a program that seems to be called
:>"traceroute" that shows the routing and speed of a connaction from thier
:>host to a remote site. But, I can't seem to find this program anywhere
:>around these parts. Is it fiction?
:>
:>What is the best way to determine the average speed of a connection between
:>my home site (crl*.crl.com) and another site?
:>
:>--
:>-Eric Penn STUPID's three rules to life:
:> stupid@crl.com 1> Stick with what you're good at,
:> 2> Learn from your mistakes,
:> 3> When in doubt, act stupid!
traceroute exists on most UNIX platforms, it basically sends a
packet to a destination but with a truncated TTL, causing an
intermediate site to return an ICMP packet. The packet is
retransmitted with a larger and larger TTL (time to live) in
order to 'trace' all the intermediate sites a packet goes through.
traceroute isn't perfect, especially when the route is asymetric.
Also, routers tend to give traceroutes and pings the lowest possible
priority.
This makes traceroute excellent at detecting congested routers, but
not so hot telling you how fast a network is or what the interactive
performance is like. Even so, the ping times are usually within 20%
of what the interactive response would be, and most people are as
interested in congestion these days as they are in interactive
response because so many backbone providers are not upgrading their
equipment to handle demand.
The only time when the ping time will be wildly off the mark from
interactive performance, at least in my experience, is when you
transit a link that performs Van-Jacobson compression... the
compression only applies to TCP packets, not ICMP/IP packets, so
ping/traceroute does not give you the actual itneractive response.
-Matt
--
Matthew Dillon director, BEST Internet Communications, Inc.
<dillon@best.com>, <dillon@apollo.west.oic.com>
[always include a portion of the original email in any response!]
Technical Details
Message-ID:
<3c5ria$h44@shell1.best.com>Path:
arc1-ca.pugleaf.net!bga.com!news.sprintlink.net!redstone.interpath.net!ddsw1!news1.best.com!shell1.best.com!not-for-mailReferences:
<3c5m64$584@crl10.crl.com>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 1120a5,a9223f802e9476af X-Google-Attributes: gid1120a5,public X-Google-Thread: fa874,a9223f802e9476af X-Google-Attributes: gidfa874,public X-Google-Thread: f4e7e,a9223f802e9476af X-Google-Attributes: gidf4e7e,public X-Google-Thread: 115608,a9223f802e9476af X-Google-Attributes: gid115608,public X-Google-ArrivalTime: 1994-12-07 23:26:52 PST Path: arc1-ca.pugleaf.net!bga.com!news.sprintlink.net!redstone.interpath.net!ddsw1!news1.best.com!shell1.best.com!not-for-mail From: dillon@best.com (Matt Dillon) Newsgroups: crl.general,ba.internet,comp.sys.amiga.datacomm,rec.games.netrek Subject: Re: Interconnect speed Date: 7 Dec 1994 18:37:45 -0801 Organization: Best Internet Communications, Inc. (info@best.com) Lines: 48 Message-ID: <3c5ria$h44@shell1.best.com> References: <3c5m64$584@crl10.crl.com> NNTP-Posting-Host: shell1.best.com