On Mon, 2008-03-10 at 18:51 -0400, Chris Snook wrote: > I suggest setting vm.overcommit_memory=2 with sysctl to turn on > explicit > overcommit control, and then try changing the value of > vm.overcommit_ratio to see if that changes where your system dies. > I've never come across these variables/commands before, so I'll go slowly here. > Also, is the system actually crashing, or does it just get swaplocked > and hang indefinitely when swap usage reaches 25%? See if you can > get > sysrq-m and sysrq-t data out of the system. You'll want to set > kernel.sysrq=1 first. Where do I find these? -- Regards Simon
