Kernel Crash when copying large files over NFS

From Linux NFS

Revision as of 15:34, 22 October 2010 by Amschuma (Talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Contents

About

  • Kernel version: 2.6.32.12, 2.6.32.16, 2.6.34.1
  • Bug 16056
  • Reported by: Philippe Dax (June 15, 2010)
  • Fixed by: Trond Myklebust (June 16, 2010)
  • Arch: amd64
  • NFS v3 client.

Symptoms

  • Copy large file to an NFS-mounted filesystem.
    • larger than the amount of available physical memory
  • swapper and rpciod on the client complain about page allocation failures.
  • kswapd goes into a deadlock, resulting in a system-wide crash.
  • Crash on UDP and TCP mounts

Cause

Resolution

This problem was fixed by:

commit b608b283a962caaa280756bc8563016a71712acf
Author: Trond Myklebust <Trond.Myklebust@netapp.com>
Date:   Fri Jul 30 15:31:54 2010 -0400

    NFS: kswapd must not block in nfs_release_page
    
    See https://bugzilla.kernel.org/show_bug.cgi?id=16056
    
    If other processes are blocked waiting for kswapd to free up some memory so
    that they can make progress, then we cannot allow kswapd to block on those
    processes.
    
    Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
    Cc: stable@kernel.org
Personal tools