[49162] in North American Network Operators' Group
Re: Testing Bandwidth performance
daemon@ATHENA.MIT.EDU (Valdis.Kletnieks@vt.edu)
Wed Jun 26 01:17:45 2002
To: Martin Hannigan <hannigan@fugawi.net>
Cc: Wojtek Zlobicki <wojtekz@idirect.com>,
Alan Sato <asato@altrio.net>, nanog@trapdoor.merit.edu
In-Reply-To: Your message of "Wed, 26 Jun 2002 01:02:29 EDT."
<5.0.2.1.0.20020626010141.014dd9a0@ns1.outland.net>
From: Valdis.Kletnieks@vt.edu
Date: Wed, 26 Jun 2002 01:16:22 -0400
Errors-To: owner-nanog-outgoing@merit.edu
--==_Exmh_-859818898P
Content-Type: text/plain; charset=us-ascii
On Wed, 26 Jun 2002 01:02:29 EDT, Martin Hannigan <hannigan@fugawi.net> said:
> Whatever happened to using NTP between sites? Works well if you have
> clocks at each site. Sorta works on strat 1/2 no clock.
That will give you latency for a specific fixed packet size, which may not be
at all correlated with actual bandwidth. You basically end up having to
assume that any jitter in the RTT is based on queues in the routers along
the way and from that extrapolate what the bandwidth was.
And of course, the first time you hit a provider that does traffic engineering
that moves NTP packets to the front of the queue so as to minimize the jitter,
the measurement becomes useless for bandwidth management...
--
Valdis Kletnieks
Computer Systems Senior Engineer
Virginia Tech
--==_Exmh_-859818898P
Content-Type: application/pgp-signature
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Exmh version 2.5 07/13/2001
iD8DBQE9GU4mcC3lWbTT17ARAqXwAKCOCQcPT+8xcesVJK+/UISJNPUYoACaA5ys
nSzaXycHsbg96irvKlUOp1E=
=Vgtj
-----END PGP SIGNATURE-----
--==_Exmh_-859818898P--
