[29726] in North American Network Operators' Group
RE: LoadBalancing products: Foundry ServerIron
daemon@ATHENA.MIT.EDU (Dmitri Krioukov)
Thu Jul 6 14:30:35 2000
From: "Dmitri Krioukov" <dima@dimension.net>
To: "Mike Diehn" <mdiehn@vicinity.com>
Cc: "Brantley Jones" <bjones@redundant.net>, <nanog@merit.edu>
Date: Thu, 6 Jul 2000 14:43:05 -0400
Message-ID: <NCBBIKACLKNMKDHKKKNFAEDIEMAA.dima@dimension.net>
MIME-Version: 1.0
Content-Type: text/plain;
charset="us-ascii"
Content-Transfer-Encoding: 7bit
In-Reply-To: <20000706115530.K13041@tupelo.vicinity.com>
Errors-To: owner-nanog-outgoing@merit.edu
as i mentioned before, if there is a
router verifying the reverse path on
the way of an ingress packet *between*
two geographically distributed lbs,
then verification may fail and the
packet is dropped.
--
dima.
> -----Original Message-----
> From: Mike Diehn [mailto:mdiehn@vicinity.com]
> Sent: Thursday, July 06, 2000 11:56 AM
> To: Dmitri Krioukov
> Cc: Brantley Jones; nanog@merit.edu
> Subject: Re: LoadBalancing products: Foundry ServerIron
>
>
> * Dmitri Krioukov (dima@dimension.net) [07 06, 2000 11:05]:
> > radware also had that rather original
> > triangle data flow mechanism i
> > mentioned. frankly speaking, i don't
> > know what happened to it. in its
> > intact form, it doesn't work anyway.
>
> Radware's triangulation works, and well.
>
> Mike
>
