On 13 Jul 2000 jmcmullan@linuxcare.com wrote:
> Rik van Riel <riel@conectiva.com.br> wrote:
> > This is a completely separate issue, but a really neat
> > idea. However, this is not something that will be able
> > to run when the system has no more memory left ;)
>
> What not at 10% (or other specific limit) swap space
> remaining - I belive that was my original suggestion,
Because it doesn't make sense. For one, if you have 1GB of
swap you don't want processes killed when you have "only"
100MB left.
Secondly, think diskless or swapless machines. In this case
the "swap full" yardstick is completely unusable since the
backing store for stuff is the filesystem and not swap (same
for mmap()ed files, etc...).
Besides, nobody will accept the fact that processes get killed
before the system runs out of memory ;)
(just watch the flamewar that will follow anyone seriously
proposing to do this)
> and means that I CAN still run. And 100k of mlock()ed
> memory would be plenty to a) renice all other processes
> to 19 and b) send a little bit of X protocol to a currenlt running
> X server or c) chvt() to an unused VT and use a text-based
> What process do you want me to kill? dialog.
OK, now envision this happening at 4AM on a "critical" mail
or webserver. You don't want the system to remain hanging until
the sysadmin walks in at 9AM...
regards,
Rik
--
"What you're running that piece of shit Gnome?!?!"
-- Miguel de Icaza, UKUUG 2000
http://www.conectiva.com/ http://www.surriel.com/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Jul 15 2000 - 21:00:17 EST