From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vivek Goyal Subject: Re: [PATCH 5/6] blkcg: move io_service_bytes and io_serviced stats into blkcg_gq Date: Fri, 26 Jun 2015 12:01:42 -0400 Message-ID: <20150626160142.GA24554@redhat.com> References: <1435268337-1738-1-git-send-email-tj@kernel.org> <1435268337-1738-6-git-send-email-tj@kernel.org> Mime-Version: 1.0 Return-path: Content-Disposition: inline In-Reply-To: <1435268337-1738-6-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org> Sender: cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Tejun Heo Cc: axboe-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, avanzini.arianna-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, kernel-team-b10kYP2dOMg@public.gmane.org On Thu, Jun 25, 2015 at 05:38:56PM -0400, Tejun Heo wrote: > Currently, both cfq-iosched and blk-throttle keep track of > io_service_bytes and io_serviced stats. While keeping track of them > separately may be useful during development, it doesn't make much > sense otherwise. Also, blkcg stats are collected at different points > and ignoring blk_do_io_stat() which can lead to subtly different > results which are more confusing than helpful. > > This patch adds ->stat_bytes and ->stat_ios to blkg (blkcg_gq), > removes the counterparts from cfq-iosched and blk-throttle and let > them print from the common blkg counters. > > The outputs are still filtered by whether the policy has > blkg_policy_data on a given blkg, so cfq's output won't show up if it > has never been used for a given blkg. The only times when the outputs > would differ significantly are when policies are attached on the fly > or elevators are switched back and forth. Those are quite exceptional > operations and I don't think they warrant keeping separate counters. > Hi Tejun, I have some questions and concerns. - Previously throttling policy was calculating number of IOs in terms of number of bios while CFQ was calculating it in terms of number of requests. I think this will be a behavior change now? IIUC, now even throttling policy will report number of requests and number of bios. - Looks like completion and stat is now lined to request. I am wondering what will happen to bio based targets. One can put a throttling policy on any of the stacked dm devices. I suspect we will be broken there? - Can you please also update the blkio-controller.txt. Specifically blkio.throttle.io_serviced. Thanks Vivek