From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bart Van Assche Subject: Re: scsi-mq Date: Fri, 13 Jun 2014 08:42:57 +0200 Message-ID: <539A9D71.7020504@acm.org> References: <1402580946-11470-1-git-send-email-hch@lst.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1402580946-11470-1-git-send-email-hch@lst.de> Sender: linux-kernel-owner@vger.kernel.org To: Christoph Hellwig , James Bottomley Cc: Jens Axboe , Robert Elliot , linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-scsi@vger.kernel.org On 06/12/14 15:48, Christoph Hellwig wrote: > The usage of blk-mq dramatically decreases CPU usage under all workloads going > down from 100% CPU usage that the old setup can hit easily to usually less > than 20% for maxing out storage subsystems with 512 byte reads and writes, > and it allows to easily achieve millions of IOPS. Bart and Robert have > helped with some very detailed measurements that they might be able to send > in reply to this, although these usually involve significantly reworked low > level drivers to avoid other bottle necks. Thanks Christoph for writing up this excellent summary. In case anyone is wondering which LLD changes Christoph is referring to: the tests I ran myself were with an SRP initiator driver modified to use multiple RDMA channels between initiator and target instead of a single channel. I will post these changes for review on the linux-rdma mailing list as soon as I have the time. Bart.