From mboxrd@z Thu Jan 1 00:00:00 1970 From: Daniel Lezcano Subject: Re: [PATCH] cgroup for disk quota Date: Fri, 20 Feb 2009 14:51:39 +0100 Message-ID: <499EB56B.6030502@free.fr> References: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org Errors-To: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org To: anqin Cc: menage-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org, containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org List-Id: containers.vger.kernel.org anqin wrote: >> from the perspective of an application developer, this approach would >> be perfect if we could have some IO bandwidth reservation mechanism >> like disk.usage_io_usage (perhaps per disk...). >> >> > > Indeed, that is my next job (exactly, current job). > > Although Paolo Valente has presented a solution of I/O bandwidth control > (see http://lwn.net/Articles/309400/), I would more like to present > bandwidth reservation mechanism, especially for data-intensive > applications (e.g. hadoop cluster). > > This kind of applications take more account of the reservation of > I/O bandwidth and netword bandwidth. > Does this feature do what you want for network bandwidth ? http://lwn.net/Articles/291161/ Option in the kernel: Networking Support -> Networking options -> QoS and/or fair queueing -> Control Group Classifier I didn't tested it yet ... :)