Too much memory causes crash when reading/writing to disk

Jeff Lessem (Jeff.Lessem@Colorado.EDU)
Mon, 16 Jul 2001 09:39:28 -0600


I have done a bit more work on the problem I reported in my message
"Crashes reading and writing to disk". To recap, on a machine with
8GB of RAM, either

dd if=/dev/zero bs=1G count=10 | split -b 1073741824

or

find /bigfulldisk -type f -exec cat {} \; > /dev/null

can reliably cause a crash. In my naive view, it seems like once a
certain amount of memory is being completely used for disk buffer
(cache, whatever, excuse my not knowing the proper term) the kernel
gets very confused and dies. The dd dies after having written about
4.5GB.

Booting the machine with mem=3000M avoids the crashes. I haven't done
a binary search to determine what the maximum amount of memory I can
use is before the problem exhibits itself. Hopefully this provides a
hint as to what could be causing the problem. If any more information
would be useful, please ask.

To go along with this, are there any parameters in /proc/sys/vm/* that
I could tune to avoid these crashes?

--
Thanks,
Jeff Lessem.
-
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/