Re: 2.6 vm/elevator loading down disks where 2.4 does not

From: Clint Byrum (cbyrum_at_spamaps.org)
Date: 06/11/04

  • Next message: Jens Axboe: "Re: [PATCH] IDE update for 2.6.7-rc3 [7/12]" 
    To: Andrew Morton <akpm@osdl.org>
Date:	Fri, 11 Jun 2004 09:45:36 -0700

    On Wed, 2004-06-09 at 16:25, Andrew Morton wrote:
    > Possibly a memory zone problem. Could you try booting with "mem=896m" on
    > the kernel command line, see how that affects things?

    This took a while longer to try, as I didn't want to unfairly test it
    against a box with 1G of RAM. So I rebooted my 2.4.23 and 2.6.7-rc3 test
    boxes with mem=896m. No change in the 5:1 ratio when comparing 2.6's
    disk reads to 2.4's. Of course, both boxes ended up reading from the
    disk more often, as they had less RAM for cache. I was unable to run a
    long test as I did before, but I'm confident the 3 hours I did run tests
    for show that this isn't a memory zone problem.

    I still think this behavior is happening because useful pages are being
    removed from the page cache too soon. Maybe this is happening because of
    excessive readahead?

    -cb

    -
    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: Jens Axboe: "Re: [PATCH] IDE update for 2.6.7-rc3 [7/12]"

    Relevant Pages

    • Re: [OT] Interview with Con Kolivas on Linux failures
      ... On one side there are people who point out that programs often allocate memory that goes unused most of the time, and this memory can be better used as disk cache. ... The ultimate result of this was the addition of a "swappiness" knob to 2.6 kernels to allow people to tune this to their taste. ... I haven't felt the need to tweak swappiness on my desktop machine from the default of 60, but it's also got 1 gigabyte of RAM. ...
      (Debian-User)
    • Re: Good settings for BlockRead
      ... > before each test to clear out the disk cache. ... with 512meg memory memory and a 2 gig file using blockread ... The interesting this is that if I force 300 mb of physical ram free ...
      (comp.lang.pascal.delphi.misc)
    • Re: Re-read an image
      ... Matlab, but geez, that gets complicated. ... This is not the same as clearing the cache. ... clear the disk cache, however I don't know how without digging into ... it's most likely not in ram memory. ...
      (comp.soft-sys.matlab)
    • Re: Re-read an image
      ... This is not the same as clearing the cache. ... way of using some low level API command to ... clear the disk cache, however I don't know how without digging into it ... the sectors gets stored in ram (I'm not certain where this ram is. ...
      (comp.soft-sys.matlab)
    • Re: Ram Drive
      ... > its cache in between that and your program, ... If it is cacheing all these temporary files, ... rattling away as the temporary files are written out to disk, ...
      (microsoft.public.windowsxp.general)
    Loading