home | help | back | first | fref | pref | prev | next | nref | lref | last | post |
Date: Fri, 19 Oct 2007 18:40:46 -0500 From: Eric Spaeth <eric@spaethco.com> Reply-To: eric@spaethco.com To: nanog@merit.edu In-Reply-To: <4719354F.8080206@labrats.us> X-SpaethCo-MailScanner-From: eric@spaethco.com Errors-To: owner-nanog@merit.edu Sean Figgins wrote: > Eric Spaeth wrote: > > > With rate-shaping they would need to have the P2P identification widget > > in-line with the data path to be able to classify and mark traffic so > > that it can be queued/throttled appropriately. > > The Sandvine, in particular, is designed to be placed in-line like > this. It does, however, deploy a technology to shunt the traffic > through the device in the event that the server craters. Many network > devices do this now. I have previous experience with Sitara QoS devices that sported that same feature. The problem was that the relay would only shut if the box lost power or if it received a software command to disengage. We had numerous problems where the packet processing engine would become overwhelmed and lock up; the relay stayed engaged because the box retained power and the software driver was rendered useless once the whole OS locked up. Maybe it's just me, but when a vendor is concerned enough about their box failing that they work out these elaborate bypass options it doesn't inspire a lot of confidence in the stability of the product. IMHO, wedging a 99.5% available piece of hardware between your 99.99+% available network hardware is just bad karma. -Eric
home | help | back | first | fref | pref | prev | next | nref | lref | last | post |