From: Randy Dunlap <email@example.com> To: 王贇 <firstname.lastname@example.org>, "Ingo Molnar" <email@example.com>, "Peter Zijlstra" <firstname.lastname@example.org>, "Juri Lelli" <email@example.com>, "Vincent Guittot" <firstname.lastname@example.org>, "Dietmar Eggemann" <email@example.com>, "Steven Rostedt" <firstname.lastname@example.org>, "Ben Segall" <email@example.com>, "Mel Gorman" <firstname.lastname@example.org>, "Luis Chamberlain" <email@example.com>, "Kees Cook" <firstname.lastname@example.org>, "Iurii Zaikin" <email@example.com>, "Michal Koutný" <firstname.lastname@example.org>, email@example.com, firstname.lastname@example.org, email@example.com, "Paul E. McKenney" <firstname.lastname@example.org>, "Jonathan Corbet" <email@example.com> Subject: Re: [PATCH v8 2/2] sched/numa: documentation for per-cgroup numa, statistics Date: Mon, 20 Jan 2020 18:08:57 -0800 Message-ID: <firstname.lastname@example.org> (raw) In-Reply-To: <email@example.com> On 1/20/20 5:57 PM, 王贇 wrote: > Add the description for 'numa_locality', also a new doc to explain > the details on how to deal with the per-cgroup numa statistics. > > Cc: Peter Zijlstra <firstname.lastname@example.org> > Cc: Michal Koutný <email@example.com> > Cc: Mel Gorman <firstname.lastname@example.org> > Cc: Jonathan Corbet <email@example.com> > Cc: Iurii Zaikin <firstname.lastname@example.org> > Cc: Randy Dunlap <email@example.com> > Signed-off-by: Michael Wang <firstname.lastname@example.org> Thanks for the updates. Acked-by: Randy Dunlap <email@example.com> > --- > Documentation/admin-guide/cg-numa-stat.rst | 178 ++++++++++++++++++++++++ > Documentation/admin-guide/index.rst | 1 + > Documentation/admin-guide/kernel-parameters.txt | 4 + > Documentation/admin-guide/sysctl/kernel.rst | 9 ++ > init/Kconfig | 2 + > 5 files changed, 194 insertions(+) > create mode 100644 Documentation/admin-guide/cg-numa-stat.rst > > diff --git a/Documentation/admin-guide/cg-numa-stat.rst b/Documentation/admin-guide/cg-numa-stat.rst > new file mode 100644 > index 000000000000..1106eb1e4050 > --- /dev/null > +++ b/Documentation/admin-guide/cg-numa-stat.rst > @@ -0,0 +1,178 @@ > +.. SPDX-License-Identifier: GPL-2.0 > + > +=============================== > +Per-cgroup NUMA statistics > +=============================== > + > +Background > +---------- > + > +On NUMA platforms, remote memory accessing always has a performance penalty. > +Although we have NUMA balancing working hard to maximize the access locality, > +there are still situations it can't help. > + > +This could happen in modern production environment. When a large number of > +cgroups are used to classify and control resources, this creates a complex > +configuration for memory policy, CPUs and NUMA nodes. In such cases NUMA > +balancing could end up with the wrong memory policy or exhausted local NUMA > +node, which would lead to low percentage of local page accesses. > + > +We need to detect such cases, figure out which workloads from which cgroup > +have introduced the issues, then we get chance to do adjustment to avoid > +performance degradation. > + > +However, there are no hardware counters for per-task local/remote accessing > +info, we don't know how many remote page accesses have occurred for a > +particular task. > + > +NUMA Locality > +------------- > + > +Fortunately, we have NUMA Balancing which scans task's mapping and triggers > +page fault periodically, giving us the opportunity to record per-task page > +accessing info, when the CPU fall into PF is from the same node of pages, we > +consider task as doing local page accessing, otherwise the remote page > +accessing, we call these two counters the locality info. > + > +On each tick, we acquire the locality info of current task on that CPU, update > +the increments into its cgroup, becoming the group locality info. > + > +By "echo 1 > /proc/sys/kernel/numa_locality" at runtime or adding boot parameter > +'numa_locality', we will enable the accounting of per-cgroup NUMA locality info, > +the 'cpu.numa_stat' entry of CPU cgroup will show statistics:: > + > + page_access local=NR_LOCAL_PAGE_ACCESS remote=NR_REMOTE_PAGE_ACCESS > + > +We define 'NUMA locality' as:: > + > + NR_LOCAL_PAGE_ACCESS * 100 / (NR_LOCAL_PAGE_ACCESS + NR_REMOTE_PAGE_ACCESS) > + > +This per-cgroup percentage number helps to represent the NUMA Balancing behavior. > + > +Note that the accounting is hierarchical, which means the NUMA locality info for > +a given group represents not only the workload of this group, but also the > +workloads of all its descendants. > + > +For example the 'cpu.numa_stat' shows:: > + > + page_access local=129909383 remote=18265810 > + > +The NUMA locality calculated as:: > + > + 129909383 * 100 / (129909383 + 18265810) = 87.67 > + > +Thus we know the workload of this group and its descendants have totally done > +129909383 times of local page accessing and 18265810 times of remotes, locality > +is 87.67% which implies most of the memory access are local. > + > +NUMA Consumption > +---------------- > + > +There are also other cgroup entries which help us to estimate NUMA efficiency. > +They are 'cpuacct.usage_percpu' and 'memory.numa_stat'. > + > +By reading 'cpuacct.usage_percpu' we will get per-cpu runtime (in nanoseconds) > +info (in hierarchy) as:: > + > + CPU_0_RUNTIME CPU_1_RUNTIME CPU_2_RUNTIME ... CPU_X_RUNTIME > + > +Combined with the info from:: > + > + cat /sys/devices/system/node/nodeX/cpulist > + > +We would be able to accumulate the runtime of CPUs into NUMA nodes, to get the > +per-cgroup node runtime info. > + > +By reading 'memory.numa_stat' we will get per-cgroup node memory consumption > +info as:: > + > + total=TOTAL_MEM N0=MEM_ON_NODE0 N1=MEM_ON_NODE1 ... NX=MEM_ON_NODEX > + > +Together we call these the per-cgroup NUMA consumption info, telling us how many > +resources a particular workload has consumed, on a particular NUMA node. > + > +Monitoring > +---------- > + > +By monitoring the increments of locality info, we can easily know whether NUMA > +Balancing is working well for a particular workload. > + > +For example we take a 5 seconds sample period, then on each sampling we have:: > + > + local_diff = last_nr_local_page_access - nr_local_page_access > + remote_diff = last_nr_remote_page_access - nr_remote_page_access > + > +and we get the locality in this period as:: > + > + locality = local_diff * 100 / (local_diff + remote_diff) > + > +We can plot a line for locality. When the line is close to 100%, things are > +good; when getting close to 0% something is wrong. We can pick a proper > +watermark to trigger warning message. > + > +You may want to drop the data if the local/remote_diff is too small, which > +implies there are not many available pages for NUMA Balancing to scan, ignoring > +would be fine since most likely the workload is insensitive to NUMA, or the > +memory topology is already good enough. > + > +Monitoring root group helps you control the overall situation, while you may > +also want to monitor all the leaf groups which contain the workloads, this > +helps to catch the mouse. > + > +Try to put your workload into also the cpuacct & memory cgroup, when NUMA > +Balancing is disabled or locality becomes too small, we may want to monitor > +the per-node runtime & memory info to see if the node consumption meet the > +requirements. > + > +For NUMA node X on each sampling we have:: > + > + runtime_X_diff = runtime_X - last_runtime_X > + runtime_all_diff = runtime_all - last_runtime_all > + > + runtime_percent_X = runtime_X_diff * 100 / runtime_all_diff > + memory_percent_X = memory_X * 100 / memory_all > + > +These two percentages are usually matched on each node, workload should execute > +mostly on the node that contains most of its memory, but it's not guaranteed. > + > +The workload may only access a small part of its memory, in such cases although > +the majority of memory are remote, locality could still be good. > + > +Thus to tell if things are fine or not depends on the understanding of system > +resource deployment, however, if you find node X got 100% memory percent but 0% > +runtime percent, definitely something is wrong. > + > +Troubleshooting > +--------------- > + > +After identifying which workload introduced the bad locality, check: > + > +1). Is the workload bound to a particular NUMA node? > +2). Has any NUMA node run out of resources? > + > +There are several ways to bind task's memory with a NUMA node, the strict way > +like the MPOL_BIND memory policy or 'cpuset.mems' will limit the memory > +node where to allocate pages. In this situation, admin should make sure the > +task is allowed to run on the CPUs of that NUMA node, and make sure there are > +available CPU resources there. > + > +There are also ways to bind task's CPU with a NUMA node, like 'cpuset.cpus' or > +sched_setaffinity() syscall. In this situation, NUMA Balancing helps to migrate > +pages into that node, admin should make sure there is available memory there. > + > +Admin could try to rebind or unbind the NUMA node to erase the damage, make a > +change then observe the statistics to see if things get better until the > +situation is acceptable. > + > +Highlights > +---------- > + > +For some tasks, NUMA Balancing may be found to be unnecessary to scan pages, > +and locality could always be 0 or small number, don't pay attention to them > +since they most likely insensitive to NUMA. > + > +There is no accounting until the option is turned on, so enable it in advance > +if you want to have the whole history. > + > +We have per-task migfailed counter to tell how many page migrations have > +failed for a particular task; you will find it in /proc/PID/sched entry.
next prev parent reply index Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top 2019-11-13 3:43 [PATCH 0/3] sched/numa: introduce advanced numa statistic 王贇 2019-11-13 3:44 ` [PATCH 1/3] sched/numa: advanced per-cgroup " 王贇 2019-11-13 3:45 ` [PATCH 2/3] sched/numa: expose per-task pages-migration-failure 王贇 2019-11-13 3:45 ` [PATCH 3/3] sched/numa: documentation for per-cgroup numa stat 王贇 2019-11-13 15:09 ` Jonathan Corbet 2019-11-14 1:52 ` 王贇 2019-11-13 18:28 ` Iurii Zaikin 2019-11-14 2:22 ` 王贇 2019-11-15 2:29 ` [PATCH v2 " 王贇 2019-11-20 9:45 ` [PATCH 0/3] sched/numa: introduce advanced numa statistic 王贇 2019-11-25 1:35 ` 王贇 2019-11-27 1:48 ` [PATCH v2 " 王贇 2019-11-27 1:49 ` [PATCH v2 1/3] sched/numa: advanced per-cgroup " 王贇 2019-11-27 10:19 ` Mel Gorman 2019-11-28 2:09 ` 王贇 2019-11-28 12:39 ` Michal Koutný 2019-11-28 13:41 ` 王贇 2019-11-28 15:58 ` Michal Koutný 2019-11-29 1:52 ` 王贇 2019-11-29 5:19 ` 王贇 2019-11-29 10:06 ` Michal Koutný 2019-12-02 2:11 ` 王贇 2019-11-27 1:50 ` [PATCH v2 2/3] sched/numa: expose per-task pages-migration-failure 王贇 2019-11-27 10:00 ` Mel Gorman 2019-12-02 2:22 ` 王贇 2019-11-27 1:50 ` [PATCH v2 3/3] sched/numa: documentation for per-cgroup numa stat 王贇 2019-11-27 4:58 ` Randy Dunlap 2019-11-27 5:54 ` 王贇 2019-12-03 5:59 ` [PATCH v3 0/2] sched/numa: introduce numa locality 王贇 2019-12-03 6:00 ` [PATCH v3 1/2] sched/numa: introduce per-cgroup NUMA locality info 王贇 2019-12-04 2:33 ` Randy Dunlap 2019-12-04 2:38 ` 王贇 2019-12-03 6:02 ` [PATCH v3 2/2] sched/numa: documentation for per-cgroup numa statistics 王贇 2019-12-03 13:43 ` Jonathan Corbet 2019-12-04 2:27 ` 王贇 2019-12-04 7:58 ` [PATCH v4 0/2] sched/numa: introduce numa locality 王贇 2019-12-04 7:59 ` [PATCH v4 1/2] sched/numa: introduce per-cgroup NUMA locality info 王贇 2019-12-05 3:28 ` Randy Dunlap 2019-12-05 3:29 ` Randy Dunlap 2019-12-05 3:52 ` 王贇 2019-12-04 8:00 ` [PATCH v4 2/2] sched/numa: documentation for per-cgroup numa statistics 王贇 2019-12-05 3:40 ` Randy Dunlap 2019-12-05 6:53 ` [PATCH v5 0/2] sched/numa: introduce numa locality 王贇 2019-12-05 6:53 ` [PATCH v5 1/2] sched/numa: introduce per-cgroup NUMA locality info 王贇 2019-12-05 6:54 ` [PATCH v5 2/2] sched/numa: documentation for per-cgroup numa, statistics 王贇 2019-12-10 2:19 ` [PATCH v5 0/2] sched/numa: introduce numa locality 王贇 2019-12-13 1:43 ` [PATCH v6 " 王贇 2019-12-13 1:47 ` [PATCH v6 1/2] sched/numa: introduce per-cgroup NUMA locality info 王贇 2020-01-03 15:14 ` Michal Koutný 2020-01-04 4:51 ` 王贇 2019-12-13 1:48 ` [PATCH v6 2/2] sched/numa: documentation for per-cgroup numa 王贇 2019-12-27 2:22 ` [PATCH v6 0/2] sched/numa: introduce numa locality 王贇 2020-01-17 2:19 ` 王贇 2020-01-19 6:08 ` [PATCH v7 " 王贇 2020-01-19 6:09 ` [PATCH v7 1/2] sched/numa: introduce per-cgroup NUMA locality info 王贇 2020-01-19 6:09 ` [PATCH v7 2/2] sched/numa: documentation for per-cgroup numa, statistics 王贇 2020-01-21 0:12 ` Randy Dunlap 2020-01-21 1:58 ` 王贇 2020-01-21 1:56 ` [PATCH v8 0/2] sched/numa: introduce numa locality 王贇 2020-01-21 1:57 ` [PATCH v8 1/2] sched/numa: introduce per-cgroup NUMA locality info 王贇 2020-01-21 1:57 ` [PATCH v8 2/2] sched/numa: documentation for per-cgroup numa, statistics 王贇 2020-01-21 2:08 ` Randy Dunlap [this message] 2020-02-07 1:10 ` [PATCH v8 0/2] sched/numa: introduce numa locality 王贇 2020-02-07 1:25 ` Steven Rostedt 2020-02-07 2:31 ` 王贇 2020-02-07 2:37 ` [PATCH RESEND " 王贇
Reply instructions: You may reply publically to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: link
Linux-Fsdevel Archive on lore.kernel.org Archives are clonable: git clone --mirror https://lore.kernel.org/linux-fsdevel/0 linux-fsdevel/git/0.git # If you have public-inbox 1.1+ installed, you may # initialize and index your mirror using the following commands: public-inbox-init -V2 linux-fsdevel linux-fsdevel/ https://lore.kernel.org/linux-fsdevel \ firstname.lastname@example.org public-inbox-index linux-fsdevel Example config snippet for mirrors Newsgroup available over NNTP: nntp://nntp.lore.kernel.org/org.kernel.vger.linux-fsdevel AGPL code for this site: git clone https://public-inbox.org/public-inbox.git