Re: swapping and the value of /proc/sys/vm/swappiness

From: Marcelo Tosatti (marcelo.tosatti_at_cyclades.com)
Date: 09/07/04

  • Next message: Dawson Engler: "[CHECKER] potential NFS deadlock in 2.6.8.1" 
    Date:	Tue, 7 Sep 2004 18:20:51 -0300
To: Con Kolivas <kernel@kolivas.org>

    On Mon, Sep 06, 2004 at 09:03:04PM -0300, Marcelo Tosatti wrote:
    > On Tue, Sep 07, 2004 at 09:34:20AM +1000, Con Kolivas wrote:
    > > Andrew Morton writes:
    > >
    > > >Con Kolivas <kernel@kolivas.org> wrote:
    > > >>
    > > >>> A scan of the change logs for swappiness related changes shows nothing
    > > >>that > might explain these changes. My question is: "Is this change in
    > > >> behavior
    > > >> > deliberate, or just a side effect of other changes that were made in
    > > >> the vm?" > and "What kind of swappiness behavior might I expect to find
    > > >> in future kernels?".
    > > >>
    > > >> The change was not deliberate but there have been some other people
    > > >> report significant changes in the swappiness behaviour as well (see
    > > >> archives). It has usually been of the increased swapping variety lately.
    > > >> It has been annoying enough to the bleeding edge desktop users for a
    > > >> swag of out-of-tree hacks to start appearing (like mine).
    > > >
    > > >All of which is largely wasted effort. It would be much more useful to get
    > > >down and identify which patch actually caused the behavioural change.
    > >
    > > I don't disagree. Is there anyone who has the time and is willing to do the
    > > regression testing? This is a general appeal to the mailing list.
    >
    > Hi kernel fellows,
    >
    > I volunteer. I'll try something tomorrow to compare swappiness of older kernels like
    > 2.6.5 and 2.6.6, which were fine on SGI's Altix tests, up to current newer kernels
    > (on small memory boxes of course).
    >
    > Someone needs to write a vmstat-like tool to parse /proc/vmstat.
    > The statistics in there allows us to watch the behaviour of VM
    > page reclaim code.
    >
    > Con, if you could compile a list of reports we would be very grateful.

    Spent some time doing a few tests.

    A contained test (512MB box, allocate 450MB of memory, touch and sleep +
    huge dd) show's that 2.6.6 and 2.6.7 are equivalent (swapout 120-150M when the "dd"
    starts writing out data).

    2.6.9-rc1 swaps out 350M on the same test. Doenst seem nice, I'll try to isolate
    what causes this tomorrow (this is probably what is hitting desktop users which
    is fixed by Con's patches). Thats problem #1. Something needs a little tweaking
    I think.

    Ray, I see the additional swapouts increase the dd performance for your particular testcase:

    on 2.6.6:
            Total I/O Avg Swap min max pg cache min max
           ----------- --------- ------- ------ --------- ------- -------
       0 242.47 MB/s 0 MB ( 0, 0) 3195 MB ( 3138, 3266)
      20 256.06 MB/s 0 MB ( 0, 0) 3170 MB ( 3074, 3234)
      40 267.29 MB/s 0 MB ( 0, 0) 3189 MB ( 3137, 3234)
      60 289.43 MB/s 666 MB ( 72, 1680) 3847 MB ( 3296, 4817) <----------

    So for this one testcase it is being beneficial.

    However, decreasing the "swappiness" value does not seem to make much of a difference:

    Kernel Version 2.6.8.1-mm4:
            Total I/O Avg Swap min max pg cache min max
           ----------- --------- ------- ------ --------- ------- -------
       0 287.28 MB/s 710 MB ( 46, 3060) 4082 MB ( 3426, 6308)
      20 288.05 MB/s 508 MB ( 94, 1417) 3848 MB ( 3442, 4739)
      40 287.03 MB/s 588 MB ( 199, 1251) 3909 MB ( 3570, 4515)
      60 290.08 MB/s 640 MB ( 210, 1190) 3976 MB ( 3538, 4531)
      80 287.73 MB/s 693 MB ( 316, 1195) 4049 MB ( 3713, 4545)
     100 166.17 MB/s 26001 MB ( 26001, 26002) 28798 MB ( 28740, 28852)
                                                                                                                                                                                       
    Kernel Version 2.6.9-rc1-mm3:
            Total I/O Avg Swap min max pg cache min max
           ----------- --------- ------- ------ --------- ------- -------
       0 274.80 MB/s 10511 MB ( 5644, 14492) 13293 MB ( 8596, 17156)
      20 267.02 MB/s 12624 MB ( 5578, 16287) 15298 MB ( 8468, 18889)
      40 267.66 MB/s 13541 MB ( 6619, 17461) 16199 MB ( 9393, 20044)
      60 233.73 MB/s 18094 MB ( 16550, 19676) 20629 MB ( 19103, 22192)
      80 213.64 MB/s 20950 MB ( 15844, 22977) 23450 MB ( 18496, 25440)
     100 164.58 MB/s 26004 MB ( 26004, 26004) 28410 MB ( 28327, 28455)

    And that is a problem #2 - swappinness not being honoured. Guys,
    any ideas on the reason for that?

    Andrew, dirty_ratio and dirty_background_ratio (as low as 5% each) did not significantly
    affect the amount of swapped out data, only a small effect on _how soon_ anonymous
    memory was swapped out.

    And finally, Ray, the difference you see between 2.6.6 and 2.6.7 can be explained,
    as noted by others in this thread, to vmscan.c changes (page replacement/scanning policy
    changes were made).

    Will continue with more tests tomorrow.
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  • Next message: Dawson Engler: "[CHECKER] potential NFS deadlock in 2.6.8.1"

    Relevant Pages

    • Re: OOM killer gripe (was Re: What still uses the block layer?)
      ... so much unless you do want the kernel to actually use it all, ... I tend to lower "swappiness" and when that happens all sorts of stuff goes ... Software suspend used to say says it can't free enough memory if I ... never triggered before hard deadlock. ...
      (Linux-Kernel)
    • swapping and the value of /proc/sys/vm/swappiness
      ... cache, and the data rates for the dd's. ... The benchmark was repeated for swappiness values of 0, 20, 40, ... What is unexpected is that the amount of swap space used at a particular ... whereas Kernel 2.6.9-rc1-mm3 swapped out 10 GB. ...
      (Linux-Kernel)
    • Re: swapping and the value of /proc/sys/vm/swappiness
      ... > swappiness setting varies dramatically with the kernel version being ... statistics' bungling or rule out statistic miscalculation as causes. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: /proc/sys/vm/swappiness == 0 makes OOM killer go beserk
      ... > tmpfs the kernel found ~250 MB of unused memory which it swapped out ... > during the last days with swappiness == 0. ... swap at all but this value decreases the swapping. ...
      (Linux-Kernel)
    • Re: [PATCH 1/3] 2.6.8-rc4-mm1 - Fix UML build
      ... can access them all, and initialized data all before uninitialized, so ... SYMLINKS:= $,$/$f) ... semaphore.c-dir = kernel ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)