Re: master 5c532fe303: Recommend that the user turn off memory overcommit

[Achim Gratz]

Po Lu writes:
> Phil Sainty <psainty@orcon.net.nz> writes:
>> I keep seeing the choice of process to kill being described as
>> "random", but it is not random.
>
> It is random, from the perspective of the user, who eventually develops
> the sinking feeling that the kernel might not kill IceCat but Emacs the
> next time it happens.

Random means different things to different people in various languages
and contexts.  From the user perspective I think I would describe the
behaviour of the OOM killer as "unpredictable".  As Phil said, if you
had the same information the kernel uses, the user could in principle
know what process it is going to select, however that data usually is
not available to the user when the OOM killer triggers.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

SD adaptations for KORG EX-800 and Poly-800MkII V0.9:
http://Synth.Stromeko.net/Downloads.html#KorgSDada