[techtalk] tty lockup

David Merrill david at lupercalia.net
Fri Mar 9 10:32:15 EST 2001


On Fri, Mar 09, 2001 at 02:32:51PM +0000, James A. Sutherland wrote:
> On Fri, 9 Mar 2001, David Merrill wrote:
> 
> > Hi,
> >
> > I was running red-carpet and it ate up ALL my available ram. I killed
> > it. Now I can get no responsiveness at the console. I am logged in
> > from another machine right now so I can send this mail.
> >
> > I suspect that the oom condition caused Linux to kill (somewhat)
> > random processes, and one of them was my tty (mgetty?) Am I on track
> > or way off base?
> >
> > Any tips on how to fix this?
> 
> A tty isn't a process, so it can't have killed that. It could have killed
> the mingetty running there, but init would have respawned it straight
> away.
> 
> The OOM killer in Linux is a fairly recent addition (partly my concept);
> until it was added, random processes would have died off; with it, the
> biggest memory hog would be killed (red-carpet in this case).

Yes, you're right. I've been following, loosely, the vm and oom
conversations in l-k (or in kernel traffic, actually) but only for the
2.4 series.

> Can you switch from one VC (virtual console) to another? Did you have the
> "Magic Alt+SysRq" support included in the kernel? If so, can you do an
> SAK? (Alt+SysRq+K - kills all processes on the current VC, which should
> give you a clean login screen.)

I cannot. I don't have the magic sysrq enabled, not being a kernel
hacker and not expecting to need it. That'll larn me.

Hmmm. I just had the thought that I might try resetting the runlevel.
Is that reasonable, or madness?

-- 
Dr. David C. Merrill                     http://www.lupercalia.net
Linux Documentation Project                   david at lupercalia.net
Collection Editor & Coordinator            http://www.linuxdoc.org
                                       Finger me for my public key

QOTD:
	"I may not be able to walk, but I drive from the sitting posistion."




More information about the Techtalk mailing list