Hi Debian,
For the first time in my (long) life oom-killer got going.
I was running 2.6.14-ck5, from here:
http://members.optusnet.com.au/ckolivas/kernel/
and it happened when I executed this *wrong* script:
echo `expr $(*blockdev --getsize* $DEVICE) \* $(blockdev --getss $DEVICE)`
I am no script whiz and all I was trying to do is figure the HD size,
copied from google.
Having met oom-killer (which did not solve anything, started swapping
and basically hung the system. How long must one wait?) I decided to
watch free memory with my monitoring apps.
I have 1GB for Ram. Free starts out as that and then gradually decreases
until it floats between 15-30MB with a full load: couple of X's, couple
of mozillas, kernel-compile, apache, mysql. But right now, not having
done much it is at 679MB. But the system has been up only half an hour.
But no oom-killer and that behavior is like that since I watched it in
2.6.12.
Has anybody seen an explanation of the oom-killer parameters for the
quick-start witted? It would be nice if you got some sort of prior
warning. But maybe that script tried to allocate everything. In this
case though it would have been better just to kill bash.
Thanks!
H
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
