[lkml]   [2010]   [Jun]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    SubjectRe: [RFC] Using "page credits" as a solution for common thrashing scenarios
    Replying to a very old email :-)

    On Wed, Nov 25, 2009 at 12:15 AM, Andi Kleen <> wrote:
    > Eyal Lotem <> writes:
    > Replying to an old email.
    >>   * I think it is wrong for the kernel to evict the 15 pages of the bash,
    >>     xterm, X server's working set, as an example, in order for a
    >>     misbehaving process to have 1000015 instead of 1000000 pages in its
    >>     working set. EVEN if that misbehaving process is accessing its working
    >>     set far more aggressively.
    > One problem in practice tends to be that it's hard to realiably detect
    > that a process is misbehaving. The 1000000 page process might be your
    > critical database, while the 15 page process is something very
    > unimportant.

    Well, this solution doesn't really depend on any detection of
    "misbehaving", it just goes about a more accurate way of defining page
    importance. A simple solution to the problem you suggest is assigning
    far more "credits" to the database than to the 15-page process.


    > -Andi
    > --
    > -- Speaking for myself only.
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to
    More majordomo info at
    Please read the FAQ at

     \ /
      Last update: 2010-06-08 11:47    [W:0.019 / U:76.772 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site