Re: [PATCH mm-unstable v1 1/4] mm/mglru: fix underprotected page cache

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Yu Zhao <yuzhao@xxxxxxxxxx> 于2023年12月15日周五 12:56写道:
>
> On Thu, Dec 14, 2023 at 04:51:00PM -0700, Yu Zhao wrote:
> > On Thu, Dec 14, 2023 at 11:38 AM Kairui Song <ryncsn@xxxxxxxxx> wrote:
> > >
> > > Yu Zhao <yuzhao@xxxxxxxxxx> 于2023年12月14日周四 11:09写道:
> > > > On Wed, Dec 13, 2023 at 12:59:14AM -0700, Yu Zhao wrote:
> > > > > On Tue, Dec 12, 2023 at 8:03 PM Kairui Song <ryncsn@xxxxxxxxx> wrote:
> > > > > >
> > > > > > Kairui Song <ryncsn@xxxxxxxxx> 于2023年12月12日周二 14:52写道:
> > > > > > >
> > > > > > > Yu Zhao <yuzhao@xxxxxxxxxx> 于2023年12月12日周二 06:07写道:
> > > > > > > >
> > > > > > > > On Fri, Dec 8, 2023 at 1:24 AM Kairui Song <ryncsn@xxxxxxxxx> wrote:
> > > > > > > > >
> > > > > > > > > Yu Zhao <yuzhao@xxxxxxxxxx> 于2023年12月8日周五 14:14写道:
> > > > > > > > > >
> > > > > > > > > > Unmapped folios accessed through file descriptors can be
> > > > > > > > > > underprotected. Those folios are added to the oldest generation based
> > > > > > > > > > on:
> > > > > > > > > > 1. The fact that they are less costly to reclaim (no need to walk the
> > > > > > > > > >    rmap and flush the TLB) and have less impact on performance (don't
> > > > > > > > > >    cause major PFs and can be non-blocking if needed again).
> > > > > > > > > > 2. The observation that they are likely to be single-use. E.g., for
> > > > > > > > > >    client use cases like Android, its apps parse configuration files
> > > > > > > > > >    and store the data in heap (anon); for server use cases like MySQL,
> > > > > > > > > >    it reads from InnoDB files and holds the cached data for tables in
> > > > > > > > > >    buffer pools (anon).
> > > > > > > > > >
> > > > > > > > > > However, the oldest generation can be very short lived, and if so, it
> > > > > > > > > > doesn't provide the PID controller with enough time to respond to a
> > > > > > > > > > surge of refaults. (Note that the PID controller uses weighted
> > > > > > > > > > refaults and those from evicted generations only take a half of the
> > > > > > > > > > whole weight.) In other words, for a short lived generation, the
> > > > > > > > > > moving average smooths out the spike quickly.
> > > > > > > > > >
> > > > > > > > > > To fix the problem:
> > > > > > > > > > 1. For folios that are already on LRU, if they can be beyond the
> > > > > > > > > >    tracking range of tiers, i.e., five accesses through file
> > > > > > > > > >    descriptors, move them to the second oldest generation to give them
> > > > > > > > > >    more time to age. (Note that tiers are used by the PID controller
> > > > > > > > > >    to statistically determine whether folios accessed multiple times
> > > > > > > > > >    through file descriptors are worth protecting.)
> > > > > > > > > > 2. When adding unmapped folios to LRU, adjust the placement of them so
> > > > > > > > > >    that they are not too close to the tail. The effect of this is
> > > > > > > > > >    similar to the above.
> > > > > > > > > >
> > > > > > > > > > On Android, launching 55 apps sequentially:
> > > > > > > > > >                            Before     After      Change
> > > > > > > > > >   workingset_refault_anon  25641024   25598972   0%
> > > > > > > > > >   workingset_refault_file  115016834  106178438  -8%
> > > > > > > > >
> > > > > > > > > Hi Yu,
> > > > > > > > >
> > > > > > > > > Thanks you for your amazing works on MGLRU.
> > > > > > > > >
> > > > > > > > > I believe this is the similar issue I was trying to resolve previously:
> > > > > > > > > https://lwn.net/Articles/945266/
> > > > > > > > > The idea is to use refault distance to decide if the page should be
> > > > > > > > > place in oldest generation or some other gen, which per my test,
> > > > > > > > > worked very well, and we have been using refault distance for MGLRU in
> > > > > > > > > multiple workloads.
> > > > > > > > >
> > > > > > > > > There are a few issues left in my previous RFC series, like anon pages
> > > > > > > > > in MGLRU shouldn't be considered, I wanted to collect feedback or test
> > > > > > > > > cases, but unfortunately it seems didn't get too much attention
> > > > > > > > > upstream.
> > > > > > > > >
> > > > > > > > > I think both this patch and my previous series are for solving the
> > > > > > > > > file pages underpertected issue, and I did a quick test using this
> > > > > > > > > series, for mongodb test, refault distance seems still a better
> > > > > > > > > solution (I'm not saying these two optimization are mutually exclusive
> > > > > > > > > though, just they do have some conflicts in implementation and solving
> > > > > > > > > similar problem):
> > > > > > > > >
> > > > > > > > > Previous result:
> > > > > > > > > ==================================================================
> > > > > > > > > Execution Results after 905 seconds
> > > > > > > > > ------------------------------------------------------------------
> > > > > > > > >                   Executed        Time (µs)       Rate
> > > > > > > > >   STOCK_LEVEL     2542            27121571486.2   0.09 txn/s
> > > > > > > > > ------------------------------------------------------------------
> > > > > > > > >   TOTAL           2542            27121571486.2   0.09 txn/s
> > > > > > > > >
> > > > > > > > > This patch:
> > > > > > > > > ==================================================================
> > > > > > > > > Execution Results after 900 seconds
> > > > > > > > > ------------------------------------------------------------------
> > > > > > > > >                   Executed        Time (µs)       Rate
> > > > > > > > >   STOCK_LEVEL     1594            27061522574.4   0.06 txn/s
> > > > > > > > > ------------------------------------------------------------------
> > > > > > > > >   TOTAL           1594            27061522574.4   0.06 txn/s
> > > > > > > > >
> > > > > > > > > Unpatched version is always around ~500.
> > > > > > > >
> > > > > > > > Thanks for the test results!
> > > > > > > >
> > > > > > > > > I think there are a few points here:
> > > > > > > > > - Refault distance make use of page shadow so it can better
> > > > > > > > > distinguish evicted pages of different access pattern (re-access
> > > > > > > > > distance).
> > > > > > > > > - Throttled refault distance can help hold part of workingset when
> > > > > > > > > memory is too small to hold the whole workingset.
> > > > > > > > >
> > > > > > > > > So maybe part of this patch and the bits of previous series can be
> > > > > > > > > combined to work better on this issue, how do you think?
> > > > > > > >
> > > > > > > > I'll try to find some time this week to look at your RFC. It'd be a
> > > > > >
> > > > > > Hi Yu,
> > > > > >
> > > > > > I'm working on V4 of the RFC now, which just update some comments, and
> > > > > > skip anon page re-activation in refault path for mglru which was not
> > > > > > very helpful, only some tiny adjustment.
> > > > > > And I found it easier to test with fio, using following test script:
> > > > > >
> > > > > > #!/bin/bash
> > > > > > swapoff -a
> > > > > >
> > > > > > modprobe brd rd_nr=1 rd_size=16777216
> > > > > > mkfs.ext4 /dev/ram0
> > > > > > mount /dev/ram0 /mnt
> > > > > >
> > > > > > mkdir -p /sys/fs/cgroup/benchmark
> > > > > > cd /sys/fs/cgroup/benchmark
> > > > > >
> > > > > > echo 4G > memory.max
> > > > > > echo $$ > cgroup.procs
> > > > > > echo 3 > /proc/sys/vm/drop_caches
> > > > > >
> > > > > > fio -name=mglru --numjobs=12 --directory=/mnt --size=1024m \
> > > > > >           --buffered=1 --ioengine=io_uring --iodepth=128 \
> > > > > >           --iodepth_batch_submit=32 --iodepth_batch_complete=32 \
> > > > > >           --rw=randread --random_distribution=zipf:0.5 --norandommap \
> > > > > >           --time_based --ramp_time=5m --runtime=5m --group_reporting
> > > > > >
> > > > > > zipf:0.5 is used here to simulate a cached read with slight bias
> > > > > > towards certain pages.
> > > > > > Unpatched 6.7-rc4:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=6548MiB/s (6866MB/s), 6548MiB/s-6548MiB/s
> > > > > > (6866MB/s-6866MB/s), io=1918GiB (2060GB), run=300001-300001msec
> > > > > >
> > > > > > Patched with RFC v4:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=7270MiB/s (7623MB/s), 7270MiB/s-7270MiB/s
> > > > > > (7623MB/s-7623MB/s), io=2130GiB (2287GB), run=300001-300001msec
> > > > > >
> > > > > > Patched with this series:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=7098MiB/s (7442MB/s), 7098MiB/s-7098MiB/s
> > > > > > (7442MB/s-7442MB/s), io=2079GiB (2233GB), run=300002-300002msec
> > > > > >
> > > > > > MGLRU off:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=6525MiB/s (6842MB/s), 6525MiB/s-6525MiB/s
> > > > > > (6842MB/s-6842MB/s), io=1912GiB (2052GB), run=300002-300002msec
> > > > > >
> > > > > > - If I change zipf:0.5 to random:
> > > > > > Unpatched 6.7-rc4:
> > > > > > Patched with this series:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=5975MiB/s (6265MB/s), 5975MiB/s-5975MiB/s
> > > > > > (6265MB/s-6265MB/s), io=1750GiB (1879GB), run=300002-300002msec
> > > > > >
> > > > > > Patched with RFC v4:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=5987MiB/s (6278MB/s), 5987MiB/s-5987MiB/s
> > > > > > (6278MB/s-6278MB/s), io=1754GiB (1883GB), run=300001-300001msec
> > > > > >
> > > > > > Patched with this series:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=5839MiB/s (6123MB/s), 5839MiB/s-5839MiB/s
> > > > > > (6123MB/s-6123MB/s), io=1711GiB (1837GB), run=300001-300001msec
> > > > > >
> > > > > > MGLRU off:
> > > > > > Run status group 0 (all jobs):
> > > > > >    READ: bw=5689MiB/s (5965MB/s), 5689MiB/s-5689MiB/s
> > > > > > (5965MB/s-5965MB/s), io=1667GiB (1790GB), run=300003-300003msec
> > > > > >
> > > > > > fio uses ramdisk so LRU accuracy will have smaller impact. The Mongodb
> > > > > > test I provided before uses a SATA SSD so it will have a much higher
> > > > > > impact. I'll provides a script to setup the test case and run it, it's
> > > > > > more complex to setup than fio since involving setting up multiple
> > > > > > replicas and auth and hundreds of GB of test fixtures, I'm currently
> > > > > > occupied by some other tasks but will try best to send them out as
> > > > > > soon as possible.
> > > > >
> > > > > Thanks! Apparently your RFC did show better IOPS with both access
> > > > > patterns, which was a surprise to me because it had higher refaults
> > > > > and usually higher refautls result in worse performance.
>
> And thanks for providing the refaults I requested for -- your data
> below confirms what I mentioned above:
>
> For fio:
>                            Your RFC   This series   Change
>   workingset_refault_file  628192729  596790506     -5%
>   IOPS                     1862k      1830k         -2%
>
> For MongoDB:
>                            Your RFC   This series   Change
>   workingset_refault_anon  10512      35277         +30%
>   workingset_refault_file  22751782   20335355      -11%
>   total                    22762294   20370632      -11%
>   TPS                      0.09       0.06          -33%
>
> For MongoDB, this series should be a big win (but apparently it's not),
> especially when using zram, since an anon refault should be a lot
> cheaper than a file refault.
>
> So, I'm baffled...
>
> One important detail I forgot to mention: based on your data from
> lru_gen_full, I think there is another difference between our Kconfigs:
>
>                   Your Kconfig  My Kconfig  Max possible
>   LRU_REFS_WIDTH  1             2           2

Hi Yu,

Thanks for the info, my fault, I forgot to update my config as I was
testing some other features.
Buf after I changed LRU_REFS_WIDTH to 2 by disabling IDLE_PAGE, thing
got much worse for MongoDB test:

With LRU_REFS_WIDTH == 2:

This patch:
==================================================================
Execution Results after 919 seconds
------------------------------------------------------------------
                  Executed        Time (µs)       Rate
  STOCK_LEVEL     488             27598136201.9   0.02 txn/s
------------------------------------------------------------------
  TOTAL           488             27598136201.9   0.02 txn/s

memcg    86 /system.slice/docker-1c3a90be9f0a072f5719332419550cd0e1455f2cd5863bc2780ca4d3f913ece5.scope
 node     0
          1     948187          0x          0x
                     0          0           0           0           0
         0           0·
                     1          0           0           0           0
         0           0·
                     2          0           0           0           0
         0           0·
                     3          0           0           0           0
         0           0·
                                0           0           0           0
         0           0·
          2     948187          0     6051788·
                     0          0r          0e          0p      11916r
     66442e          0p
                     1          0r          0e          0p        903r
     16888e          0p
                     2          0r          0e          0p        459r
      9764e          0p
                     3          0r          0e          0p          0r
         0e       2874p
                                0           0           0           0
         0           0·
          3     948187    1353160        6351·
                     0          0           0           0           0
         0           0·
                     1          0           0           0           0
         0           0·
                     2          0           0           0           0
         0           0·
                     3          0           0           0           0
         0           0·
                                0           0           0           0
         0           0·
          4      73045      23573          12·
                     0          0R          0T          0     3498607R
   4868605T          0·
                     1          0R          0T          0     3012246R
   3270261T          0·
                     2          0R          0T          0     2498608R
   2839104T          0·
                     3          0R          0T          0           0R
   1983947T          0·
                          1486579L          0O    1380614Y       2945N
      2945F       2734A

workingset_refault_anon 0
workingset_refault_file 18130598

              total        used        free      shared  buff/cache   available
Mem:          31978        6705         312          20       24960       24786
Swap:         31977           4       31973

RFC:
==================================================================
Execution Results after 908 seconds
------------------------------------------------------------------
                  Executed        Time (µs)       Rate
  STOCK_LEVEL     2252            27159962888.2   0.08 txn/s
------------------------------------------------------------------
  TOTAL           2252            27159962888.2   0.08 txn/s

workingset_refault_anon 22585
workingset_refault_file 22715256

memcg    66 /system.slice/docker-0989446ff78106e32d3f400a0cf371c9a703281bded86d6d6bb1af706ebb25da.scope
 node     0
         22     563007       2274     1198225·
                     0          0r          1e          0p          0r
    697076e          0p
                     1          0r          0e          0p          0r
         0e     325661p
                     2          0r          0e          0p          0r
         0e     888728p
                     3          0r          0e          0p          0r
         0e    3602238p
                                0           0           0           0
         0           0·
         23     532222       7525     4948747·
                     0          0           0           0           0
         0           0·
                     1          0           0           0           0
         0           0·
                     2          0           0           0           0
         0           0·
                     3          0           0           0           0
         0           0·
                                0           0           0           0
         0           0·
         24     500367    1214667        3292·
                     0          0           0           0           0
         0           0·
                     1          0           0           0           0
         0           0·
                     2          0           0           0           0
         0           0·
                     3          0           0           0           0
         0           0·
                                0           0           0           0
         0           0·
         25     469692      40797         466·
                     0          0R        271T          0           0R
   1162165T          0·
                     1          0R          0T          0      774028R
   1205332T          0·
                     2          0R          0T          0           0R
    932484T          0·
                     3          0R          1T          0           0R
   4252158T          0·
                         25178380L     156515O   23953602Y      59234N
     49391F      48664A

              total        used        free      shared  buff/cache   available
Mem:          31978        6968         338           5       24671       24555
Swap:         31977        1533       30444

Using same mongodb config (a 3 replica cluster using the same config):
{
    "net": {
        "bindIpAll": true,
        "ipv6": false,
        "maxIncomingConnections": 10000,
    },
    "setParameter": {
        "disabledSecureAllocatorDomains": "*"
    },
    "replication": {
        "oplogSizeMB": 10480,
        "replSetName": "issa-tpcc_0"
    },
    "security": {
        "keyFile": "/data/db/keyfile"
    },
    "storage": {
        "dbPath": "/data/db/",
        "syncPeriodSecs": 60,
        "directoryPerDB": true,
        "wiredTiger": {
            "engineConfig": {
                "cacheSizeGB": 5
            }
        }
    },
    "systemLog": {
        "destination": "file",
        "logAppend": true,
        "logRotate": "rename",
        "path": "/data/db/mongod.log",
        "verbosity": 0
    }
}

The test environment have 32g memory and 16 core.

Per my analyze, the access pattern for the mongodb test is that page
will be re-access long after it's evicted so PID controller won't
protect higher tier. That RFC will make use of the long existing
shadow to do feedback to PID/Gen so the result will be much better.
Still need more adjusting though, will try to do a rebase on top of
mm-unstable which includes your patch.

I've no idea why the workingset_refault_* is higher in the better
case, this a clearly an IO bound workload, Memory and IO is busy while
CPU is not full...

I've uploaded my local reproducer here:
https://github.com/ryncsn/emm-test-project/tree/master/mongo-cluster
https://github.com/ryncsn/py-tpcc





[Index of Archives]     [Linux Kernel]     [Kernel Development Newbies]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Hiking]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux