Message-ID: <a1262b55-c6a1-d3a6-6715-61ebcfad9f9f@linux.alibaba.com>
Date:   Tue, 23 Apr 2019 10:14:38 +0800
From:   王贇 <yun.wang@...ux.alibaba.com>
To:     禹舟键 <ufo19890607@...il.com>
Cc:     Peter Zijlstra <peterz@...radead.org>, hannes@...xchg.org,
        mhocko@...nel.org, vdavydov.dev@...il.com,
        Ingo Molnar <mingo@...hat.com>, linux-kernel@...r.kernel.org,
        linux-mm@...ck.org
Subject: Re: [RFC PATCH 0/5] NUMA Balancer Suite

On 2019/4/22 下午10:34, 禹舟键 wrote:
> Hi, Michael
> I really want to know how could you fix the conflict between numa balancer and load balancer. Maybe you gained numa bonus by migrating some tasks to the node with most of the cache there, but, cpu load balance was break, so how to do it ?

The trick here is to allow migration when load balancing keep failing,
which means no better tasks to move.

However, since the idea here is cgroup workloads scheduling, it could be
hard to make sure load balanced, for example only two cgroup with different
workloads and putting them to different node.

Thus why we make this a module, rather than changing the kernel logical,
at this moment not every situation could gain benefit from numa balancer,
but in some situations, balanced load can't bring benefit while numa
balancer could.

Also we are improving the module to give it an overall sight, so it will
know whether the decision is breaking the load balance, but this introduced
big lock and more per cpu/node counters, we need more testing to know whether
this is really helpful.

Anyway, if you have any scenery may could gain benefit, please take a try
and let me know what's the problem is, we'll try to address them :-)

Regards,
Michael Wang

> 
> Thanks
> Wind
> 
> 
> 王贇 <yun.wang@...ux.alibaba.com <mailto:yun.wang@...ux.alibaba.com>> 于2019年4月22日周一 上午10:13写道：
> 
>     We have NUMA Balancing feature which always trying to move pages
>     of a task to the node it executed more, while still got issues:
> 
>     * page cache can't be handled
>     * no cgroup level balancing
> 
>     Suppose we have a box with 4 cpu, two cgroup A & B each running 4 tasks,
>     below scenery could be easily observed:
> 
>     NODE0                   |       NODE1
>                             |
>     CPU0            CPU1    |       CPU2            CPU3
>     task_A0         task_A1 |       task_A2         task_A3
>     task_B0         task_B1 |       task_B2         task_B3
> 
>     and usually with the equal memory consumption on each node, when tasks have
>     similar behavior.
> 
>     In this case numa balancing try to move pages of task_A0,1 & task_B0,1 to node 0,
>     pages of task_A2,3 & task_B2,3 to node 1, but page cache will be located randomly,
>     depends on the first read/write CPU location.
> 
>     Let's suppose another scenery:
> 
>     NODE0                   |       NODE1
>                             |
>     CPU0            CPU1    |       CPU2            CPU3
>     task_A0         task_A1 |       task_B0         task_B1
>     task_A2         task_A3 |       task_B2         task_B3
> 
>     By switching the cpu & memory resources of task_A0,1 and task_B0,1, now workloads
>     of cgroup A all on node 0, and cgroup B all on node 1, resource consumption are same
>     but related tasks could share a closer cpu cache, while cache still randomly located.
> 
>     Now what if the workloads generate lot's of page cache, and most of the memory
>     accessing are page cache writing?
> 
>     A page cache generated by task_A0 on NODE1 won't follow it to NODE0, but if task_A0
>     was already on NODE0 before it read/write files, caches will be there, so how to
>     make sure this happen?
> 
>     Usually we could solve this problem by binding workloads on a single node, if the
>     cgroup A was binding to CPU0,1, then all the caches it generated will be on NODE0,
>     the numa bonus will be maximum.
> 
>     However, this require a very well administration on specified workloads, suppose in our
>     cases if A & B are with a changing CPU requirement from 0% to 400%, then binding to a
>     single node would be a bad idea.
> 
>     So what we need is a way to detect memory topology on cgroup level, and try to migrate
>     cpu/mem resources to the node with most of the caches there, as long as the resource
>     is plenty on that node.
> 
>     This patch set introduced:
>       * advanced per-cgroup numa statistic
>       * numa preferred node feature
>       * Numa Balancer module
> 
>     Which helps to achieve an easy and flexible numa resource assignment, to gain numa bonus
>     as much as possible.
> 
>     Michael Wang (5):
>       numa: introduce per-cgroup numa balancing locality statistic
>       numa: append per-node execution info in memory.numa_stat
>       numa: introduce per-cgroup preferred numa node
>       numa: introduce numa balancer infrastructure
>       numa: numa balancer
> 
>      drivers/Makefile             |   1 +
>      drivers/numa/Makefile        |   1 +
>      drivers/numa/numa_balancer.c | 715 +++++++++++++++++++++++++++++++++++++++++++
>      include/linux/memcontrol.h   |  99 ++++++
>      include/linux/sched.h        |   9 +-
>      kernel/sched/debug.c         |   8 +
>      kernel/sched/fair.c          |  41 +++
>      mm/huge_memory.c             |   7 +-
>      mm/memcontrol.c              | 246 +++++++++++++++
>      mm/memory.c                  |   9 +-
>      mm/mempolicy.c               |   4 +
>      11 files changed, 1133 insertions(+), 7 deletions(-)
>      create mode 100644 drivers/numa/Makefile
>      create mode 100644 drivers/numa/numa_balancer.c
> 
>     -- 
>     2.14.4.44.g2045bb6
> 

Powered by blists - more mailing lists