From: Michal Hocko Subject: Re: [PATCH v2 0/5] Introduce memcg_stock_pcp remote draining Date: Fri, 27 Jan 2023 07:58:32 +0100 Message-ID: References: <20230125073502.743446-1-leobras@redhat.com> <9e61ab53e1419a144f774b95230b789244895424.camel@redhat.com> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.com; s=susede1; t=1674802714; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=g+E5xBJ3UOLpBqu7Q/7w5cQuQLqZh3v0sZZi4etorHM=; b=YGGNoTNQSsBG2crAEEf577da9ZlMzCQfCzd1kLQO55iHcOKNihErXr7ZcEAiMcI3Nhh4OP Po3w6eO0WJ6HNVdNfx5oWkuZax3loRjgY6iHwTSCaqhEVGvHYBknYbOSpV4VkJ4h2CvyAp V4rM4uxPXKoxjeLA7/drrCxbfEz8Wb4= Content-Disposition: inline In-Reply-To: List-ID: Content-Type: text/plain; charset="iso-8859-1" To: Marcelo Tosatti Cc: Roman Gushchin , Leonardo =?iso-8859-1?Q?Br=E1s?= , Johannes Weiner , Shakeel Butt , Muchun Song , Andrew Morton , cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org On Thu 26-01-23 21:32:18, Marcelo Tosatti wrote: > On Thu, Jan 26, 2023 at 08:20:46PM +0100, Michal Hocko wrote: > > On Thu 26-01-23 15:03:43, Marcelo Tosatti wrote: > > > On Thu, Jan 26, 2023 at 08:41:34AM +0100, Michal Hocko wrote: > > > > On Wed 25-01-23 15:14:48, Roman Gushchin wrote: > > > > > On Wed, Jan 25, 2023 at 03:22:00PM -0300, Marcelo Tosatti wrote: > > > > > > On Wed, Jan 25, 2023 at 08:06:46AM -0300, Leonardo Br=E1s wrote: > > > > > > > On Wed, 2023-01-25 at 09:33 +0100, Michal Hocko wrote: > > > > > > > > On Wed 25-01-23 04:34:57, Leonardo Bras wrote: > > > > > > > > > Disclaimer: > > > > > > > > > a - The cover letter got bigger than expected, so I had t= o split it in > > > > > > > > > sections to better organize myself. I am not very con= fortable with it. > > > > > > > > > b - Performance numbers below did not include patch 5/5 (= Remove flags > > > > > > > > > from memcg_stock_pcp), which could further improve pe= rformance for > > > > > > > > > drain_all_stock(), but I could only notice the optimi= zation at the > > > > > > > > > last minute. > > > > > > > > >=20 > > > > > > > > >=20 > > > > > > > > > 0 - Motivation: > > > > > > > > > On current codebase, when drain_all_stock() is ran, it wi= ll schedule a > > > > > > > > > drain_local_stock() for each cpu that has a percpu stock = associated with a > > > > > > > > > descendant of a given root_memcg. > > > > >=20 > > > > > Do you know what caused those drain_all_stock() calls? I wonder i= f we should look > > > > > into why we have many of them and whether we really need them? > > > > >=20 > > > > > It's either some user's actions (e.g. reducing memory.max), eithe= r some memcg > > > > > is entering pre-oom conditions. In the latter case a lot of drain= calls can be > > > > > scheduled without a good reason (assuming the cgroup contain mult= iple tasks running > > > > > on multiple cpus). > > > >=20 > > > > I believe I've never got a specific answer to that. We > > > > have discussed that in the previous version submission > > > > (20221102020243.522358-1-leobras-H+wXaHxf7aLQT0dZR+AlfA@public.gman= e.org and specifically > > > > Y2TQLavnLVd4qHMT-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org). Leonardo= has mentioned a mix of RT and > > > > isolcpus. I was wondering about using memcgs in RT workloads because > > > > that just sounds weird but let's say this is the case indeed.=20 > > >=20 > > > This could be the case. You can consider an "edge device" where it is > > > necessary to run a RT workload. It might also be useful to run=20 > > > non realtime applications on the same system. > > >=20 > > > > Then an RT task or whatever task that is running on an isolated > > > > cpu can have pcp charges. > > >=20 > > > Usually the RT task (or more specifically the realtime sensitive loop > > > of the application) runs entirely on userspace. But i suppose there > > > could be charges on application startup. > >=20 > > What is the role of memcg then? If the memory limit is in place and the > > workload doesn't fit in then it will get reclaimed during start up and > > memory would need to be refaulted if not mlocked. If it is mlocked then > > the limit cannot be enforced and the start up would likely fail as a > > result of the memcg oom killer. >=20 > 1) Application which is not time sensitive executes on isolated CPU, > with memcg control enabled. Per-CPU stock is created. >=20 > 2) App with memcg control enabled exits, per-CPU stock is not drained. >=20 > 3) Latency sensitive application starts, isolated per-CPU has stock to > be drained, and: >=20 > /* > * Drains all per-CPU charge caches for given root_memcg resp. subtree > * of the hierarchy under it. > */ > static void drain_all_stock(struct mem_cgroup *root_memcg) No, this is not really answering my question. See Y9LQ615H13RmG7wL-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org which already expl= ains how the draining would be triggered. This is not really happening on any operation. I am really asking for specific workloads which are running multiple processes on a mix of isolated and non-isolated cpus yet they share memcg so that they can interfere. The consequences of the common memcg are described above. --=20 Michal Hocko SUSE Labs