From mboxrd@z Thu Jan 1 00:00:00 1970 From: Benjamin LaHaise Subject: Re: [PATCH 1/3] sysfs directory scaling: rbtree for dirent name lookups Date: Tue, 3 Nov 2009 11:45:50 -0500 Message-ID: <20091103164550.GM8227@kvack.org> References: <20091101163130.GA7911@kvack.org> <20091103035058.GA19515@kroah.com> <4AEFCA49.4020305@gmail.com> <20091103160715.GD23857@kroah.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Eric Dumazet , "Eric W. Biederman" , Octavian Purdila , netdev@vger.kernel.org, Cosmin Ratiu , linux-kernel@vger.kernel.org To: Greg KH Return-path: Received: from kanga.kvack.org ([205.233.56.17]:49759 "EHLO kanga.kvack.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750879AbZKCQpq (ORCPT ); Tue, 3 Nov 2009 11:45:46 -0500 Content-Disposition: inline In-Reply-To: <20091103160715.GD23857@kroah.com> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, Nov 03, 2009 at 08:07:15AM -0800, Greg KH wrote: > But registering 20000 devices is a far different problem from using > those 20000 devices :) Registering 20,000 devices *is* a real world problem (I'm actually aiming for 100,000, as that's what roughly fits in a single 10Gbps link -- something that a mid range system can now route). When an edge router comes up from reboot, or after a link has been down, the rate at which customers connect is important -- too slow, and you get a pile of support calls from customers complaining that their connection is down. Because of the data structures used, there isn't even any improvement from an SMP system, so this needs to be addressed directly. -ben