[192845] in North American Network Operators' Group
Re: 10G switch drops traffic for a split second
daemon@ATHENA.MIT.EDU (John Kristoff)
Tue Nov 29 12:39:45 2016
X-Original-To: nanog@nanog.org
Date: Tue, 29 Nov 2016 11:39:41 -0600
From: John Kristoff <jtk@depaul.edu>
To: TJ Trout <tj@pcguys.us>
In-Reply-To: <e3b6a273e05e4b958b0e4a652f04964b@XCASPRD01-DFT.dpu.depaul.edu>
Reply-To: jtk@aharp.iorc.depaul.edu
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
On Tue, 29 Nov 2016 09:06:00 +0000
TJ Trout <tj@pcguys.us> wrote:
> Could this be MTU? I've tried flow control, hard code duplex, stp on/off etc
> I'm at a loss any ideas?
This sounds like a common problem that certain data center environments
run into with 10 Gb/s and higher loads. In a nutshell, a simple
throw money at it solution to the problem is to use a switch with larger
buffers.
Nonetheless, there are plenty of good sources of material about the
phenomenon, such as:
<http://fasterdata.es.net/network-tuning/router-switch-buffer-size-issues/>
<https://people.ucsc.edu/~warner/buffer.html>
John
