Date:   Tue, 1 Mar 2022 15:51:22 +0800
From:   Miaohe Lin <linmiaohe@...wei.com>
To:     "Huang, Ying" <ying.huang@...el.com>
CC:     <linux-mm@...ck.org>, <linux-kernel@...r.kernel.org>,
        Feng Tang <feng.tang@...el.com>,
        Baolin Wang <baolin.wang@...ux.alibaba.com>,
        Michal Hocko <mhocko@...e.com>,
        Rik van Riel <riel@...riel.com>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        Yang Shi <shy828301@...il.com>, Zi Yan <ziy@...dia.com>,
        Wei Xu <weixugc@...gle.com>,
        Oscar Salvador <osalvador@...e.de>,
        Shakeel Butt <shakeelb@...gle.com>,
        zhongjiang-ali <zhongjiang-ali@...ux.alibaba.com>,
        Randy Dunlap <rdunlap@...radead.org>,
        Johannes Weiner <hannes@...xchg.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Mel Gorman <mgorman@...hsingularity.net>,
        Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH -V13 2/3] NUMA balancing: optimize page placement for
 memory tiering system

On 2022/3/1 14:47, Huang, Ying wrote:
> Miaohe Lin <linmiaohe@...wei.com> writes:
> 
>> On 2022/2/21 16:45, Huang Ying wrote:
>>> With the advent of various new memory types, some machines will have
>>> multiple types of memory, e.g. DRAM and PMEM (persistent memory).  The
>>> memory subsystem of these machines can be called memory tiering
>>> system, because the performance of the different types of memory are
>>> usually different.
>>>
>>> In such system, because of the memory accessing pattern changing etc,
>>> some pages in the slow memory may become hot globally.  So in this
>>> patch, the NUMA balancing mechanism is enhanced to optimize the page
>>> placement among the different memory types according to hot/cold
>>> dynamically.
>>>
>>> In a typical memory tiering system, there are CPUs, fast memory and
>>> slow memory in each physical NUMA node.  The CPUs and the fast memory
>>> will be put in one logical node (called fast memory node), while the
>>> slow memory will be put in another (faked) logical node (called slow
>>> memory node).  That is, the fast memory is regarded as local while the
>>> slow memory is regarded as remote.  So it's possible for the recently
>>> accessed pages in the slow memory node to be promoted to the fast
>>> memory node via the existing NUMA balancing mechanism.
>>>
>>> The original NUMA balancing mechanism will stop to migrate pages if
>>> the free memory of the target node becomes below the high watermark.
>>> This is a reasonable policy if there's only one memory type.  But this
>>> makes the original NUMA balancing mechanism almost do not work to
>>> optimize page placement among different memory types.  Details are as
>>> follows.
>>>
>>> It's the common cases that the working-set size of the workload is
>>> larger than the size of the fast memory nodes.  Otherwise, it's
>>> unnecessary to use the slow memory at all.  So, there are almost
>>> always no enough free pages in the fast memory nodes, so that the
>>> globally hot pages in the slow memory node cannot be promoted to the
>>> fast memory node.  To solve the issue, we have 2 choices as follows,
>>>
>>> a. Ignore the free pages watermark checking when promoting hot pages
>>>    from the slow memory node to the fast memory node.  This will
>>>    create some memory pressure in the fast memory node, thus trigger
>>>    the memory reclaiming.  So that, the cold pages in the fast memory
>>>    node will be demoted to the slow memory node.
>>>
>>> b. Make kswapd of the fast memory node to reclaim pages until the free
>>>    pages are a little more than the high watermark (named as promo
>>>    watermark).  Then, if the free pages of the fast memory node reaches
>>>    high watermark, and some hot pages need to be promoted, kswapd of the
>>>    fast memory node will be waken up to demote more cold pages in the
>>>    fast memory node to the slow memory node.  This will free some extra
>>>    space in the fast memory node, so the hot pages in the slow memory
>>>    node can be promoted to the fast memory node.
>>>
>>> The choice "a" may create high memory pressure in the fast memory
>>> node.  If the memory pressure of the workload is high, the memory
>>> pressure may become so high that the memory allocation latency of the
>>> workload is influenced, e.g. the direct reclaiming may be triggered.
>>>
>>> The choice "b" works much better at this aspect.  If the memory
>>> pressure of the workload is high, the hot pages promotion will stop
>>> earlier because its allocation watermark is higher than that of the
>>
>> Many thanks for your path. The patch looks good to me but I have a question.
>> WMARK_PROMO is only used inside pgdat_balanced when NUMA_BALANCING_MEMORY_TIERING
>> is set. So its allocation watermark seems to be as same as the normal memory
>> allocation. How should I understand the above sentence? Am I miss something?
> 
> Before allocating pages for promotion, the watermark of the fast node
> will be checked (please refer to migrate_balanced_pgdat()).  If the
> watermark is going to be lower than the high watermark, promotion will
> abort.

I see. The hot pages promotion watermark is "nr_migrate_pages" more than that of the
normal memory allocation not "_watermark[WMARK_PROMO] - _watermark[WMARK_HIGH]".

Many thanks for your kindly explanation. :)

> 
> Best Regards,
> Huang, Ying
> .
> 

Powered by blists - more mailing lists