[87799] in North American Network Operators' Group
Re: workhorse of the future...
daemon@ATHENA.MIT.EDU (Richard A Steenbergen)
Wed Jan 11 01:48:22 2006
Date: Wed, 11 Jan 2006 01:47:48 -0500
From: Richard A Steenbergen <ras@e-gerbil.net>
To: bmanning@vacation.karoshi.com
Cc: nanog@nanog.org
In-Reply-To: <20060110223959.GA28449@vacation.karoshi.com.>
Errors-To: owner-nanog@merit.edu
On Tue, Jan 10, 2006 at 10:39:59PM +0000, bmanning@vacation.karoshi.com wro=
te:
>=20
> first it was the vitalinks, then the bridge gear, then proteon, then cisc=
o AGS,
> then 7600VXR, then 7301s....
>=20
> looking to find the next-gen workhorse ... looking for 4-6yr life expecta=
ncy.
> pointers(private are ok) are appreciated - as well as -why- you think the
> suggested boxen are likely candidates.
You know if I didn't know better, I would think this was a troll. :)
Everyone's workhorses are different, depending on what kind of work you=20
want to do with them. The 7200VXR (which is what I assume you meant) and=20
7301 are the last great "mostly cpu based" routers out there, which is why=
=20
they have lasted as long and become as widely used as they have. Any time=
=20
you have a CPU based solution it is going to be easier to add new features=
=20
quickly, and handle a wide variety of tasks. Personally I couldn't find a=
=20
use for either one in my network on a dare, but that is because I care=20
about capacity not high touch services or L2TP termination.
It is pretty hard to predict what box is going to be the "it" thing for=20
the next 5 years, though I certainly agree that anyone interested in=20
making smart purchases needs to be concerned about the viability of the=20
product over exactly that kind of timeframe. For my needs, the Juniper=20
M160 has been the workhorse for the past 3-5 years, though its time is=20
rapidly coming to an end. The weapon of choice for L3 ethernet aggregation=
=20
is certainly now the 6500/7600 SUP720 based platforms, and will probably=20
see a good 5 years worth of use and deployment. Folks like Foundry,=20
Extreme, and Force10 have all come out with interesting "nextgen" boxes at=
=20
L2, but I think they've already lost the L3 war to Cisco before firing a=20
single shot.
I'm not entirely certain that the next 5 years has been ironed out in the=
=20
carrier space yet. There is still plenty of opportunity for Juniper to be=
=20
dethroned if they follow through with some of the disturbing trends=20
they've been setting (and from all indications, we won't be seeing=20
anything new or even close to revolutionary for at least 2 years). The=20
point has been made that this pattern of becoming complacent re:=20
innovation and cost effectiveness until you get your ass handed to you by=
=20
a competetive product is the natural cycle of things, and we may very well=
=20
be near another turning point in the market like what happened when=20
Juniper first hit the scene. The CRS1 hasn't made significant headway into=
=20
the market yet either though, most likely due to its lack of any low-end=20
or "non-40Gbps" cards as an upgrade path for existing GSR users.
I wouldn't be surprised to see the 7206VXR and 7301 still in use 5 years=20
=66rom now though, some roles just aren't that demanding traffic-wise, and=
=20
are better served by a CPU based solution. I couldn't tell you if there=20
are plans for a bigger beefier CPU-based router (not my area of concern=20
really), but in that space I think the 7300 may be a safe bet for the near=
=20
future.
--=20
Richard A Steenbergen <ras@e-gerbil.net> http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)
