[144517] in North American Network Operators' Group
Re: vyatta for bgp
daemon@ATHENA.MIT.EDU (Robert Bays)
Mon Sep 12 21:04:44 2011
From: Robert Bays <robert@vyatta.com>
In-Reply-To: <C7EE963D-5836-43B1-8378-5D194F562FE0@vyatta.com>
Date: Mon, 12 Sep 2011 18:03:59 -0700
To: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
> On Sep 13, 2011, at 2:45 AM, Roland Dobbins wrote:
> This contradicts my experience - I've repeatedly witnessed only a few =
mb/sec of 64-byte packets making software-based routers fall over, =
including just last month.
It's easy to get 6Mpps using Vyatta or most other software based routers =
on modern server grade hardware. If both your link and hardware =
capacity don't exceed the maximum load the bad guys can throw at you, =
then you are doomed to failure regardless without upstream cooperation. =
In the case of an enterprise, neither of those two things are always =
guaranteed. Software based routers work great for many use cases, it =
just depends on what you are looking for. And sometimes they are the =
only option, for example as services migrate into the cloud.
Software based routers are only going to get better. Roland, if you are =
in the Bay Area I invite you to drop by Vyatta HQ in Belmont and I will =
show you a demo of our next generation software which is doing an 5x =
increase in throughput over our last generation on the same off the =
shelf Intel X86 processors. We expect that number to only increase over =
the short term. Lot's of CPU horsepower solves a multitude of woes.
The one thing I do agree with you on is=85
> One can and should test the specific performance envelope of any =
prospective infrastructure purchase, of course.
cheers,
robert.=
