Re: page allocation failures in mv643xx_eth_poll (was Re: Swapper problems?)
-----BEGIN PGP SIGNED MESSAGE-----
Lennert Buytenhek wrote:
> The kernel can end up killing processes if there is not enough memory
> to satisfy a memory allocation request made by one of the processes in
> the system.
Yes, I'm aware of that --- I've been chasing this thing up for years.
> If you're seeing both temporary memory allocation failures and killed
> processes, you're seeing something else than the original poster is.
Obviously, although the messages I was getting are extremely similar.
There have been lots of complaints about the OOM killer kicking in and
spuriously killing stuff on the NSLU2 list over the years; I've seen
some signs ---
http://lkml.indiana.edu/hypermail/linux/kernel/0711.3/0766.html --- that
it's related to using USB storage. However I've never actually seen
anyone come up with a proper diagnosis. I would love to know if this has
been properly fixed. It's not so much of a problem on my SheevaPlug, but
it was making some operations impossible on the NSLU2.
As I said, turning overcommit off appears to help on my SheevaPlug, but
that could just be due to changing usage patterns. I haven't noticed any
services mysteriously shut down since I added the setting, but before
that my Java server was dying on a regular basis.
(PS. Please don't cc me; I'm on the list, I don't need two copies!)
┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ───── http://www.cowlark.com ─────
│ "Under communism, man exploits man. Under capitalism, it's just the
│ opposite." --- John Kenneth Galbrith
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
-----END PGP SIGNATURE-----