On Mon, 17.09.2007 at 20:04:24 -0400, Michael Stone wrote:
> I'm going to reassign this to linux-image. The basic problem seems to be
> the loss of responsiveness, which is probably caused by excessive
> amounts of large block size IO on the encrypted volume. There's a
> possible issue with the OOM killer.

FWIW, today I also saw OOM killer in (broken?) action. Imho, OOM killer
should only trigger if there's at least either no more RAM to work
with, and note this in the syslog, or when swap is exhausted.

My kern.log has this, with the time stamp and hostname removed:

DMA: 465*4kB 17*8kB 1*16kB 1*32kB 1*64kB 0*128kB 1*256kB 1*512kB 1*1024kB 1*2048kB 0*4096kB = 5948kB
DMA32: empty
Normal: empty
HighMem: empty
Swap cache: add 2195725, delete 2046650, find 782418/975703, race 1+47
Free swap = 2489952kB
Total swap = 4194296kB
Free swap: 2489952kB
483649 pages of RAM
14999 reserved pages
70 pages shared
149071 pages swap cached

Unfortunately, I can't make too much out of these numbers.


To UNSUBSCRIBE, email to debian-bugs-dist-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org