Re: Linux killed Kenny, bastard!

Re: Linux killed Kenny, bastard!

[Tomasz Chmielewski]

>> On Tue, 13 Jan 2009, Evgeniy Polyakov wrote:
>> 
>> > Like anything that spawns a thread or process per request/client, or
>> > preallocates set of them which connect to the huge object like database.
>> > Most of the time database/server is killed first instead of comparably
>> > small clients.
>> 
>> No, the reverse is true: when a task is chosen for oom kill based on the 
>> badness heuristic, the oom killer first attempts to kill any child task 
>> that isn't attached to the same mm.  If the child shares an mm, both tasks 
>> must die before memory freeing can occur.
> 
> It is a theory, not a practice. OOM-killer most of time starts from ssh,
> database and lighttpd on the tested machines, when it could start in
> the reverse order and do not touch ssh at all. Better not from daemon
> itself, but its fastcgi spawned processes.

How does this feature relate to:

config ANDROID_LOW_MEMORY_KILLER
         bool "Android Low Memory Killer"
         default N
         ---help---
           Register processes to be killed when memory is low

available in Staging drivers / Android?


-- 
Tomasz Chmielewski
http://wpkg.org

ANDROID low memory killer register (was Linux killed Kenny, bastard!)

[Alan Cox]

> config ANDROID_LOW_MEMORY_KILLER
>          bool "Android Low Memory Killer"
>          default N
>          ---help---
>            Register processes to be killed when memory is low
> 
> available in Staging drivers / Android?

I believe they are related in the sense that both of them are
unneccessary and should not make the final kernel...

The /proc../oom_adj interface is again sufficient for this.

Re: Linux killed Kenny, bastard!

[Evgeniy Polyakov]

On Tue, Jan 13, 2009 at 11:58:38AM +0100, Tomasz Chmielewski (mangoo@wpkg.org) wrote:
> config ANDROID_LOW_MEMORY_KILLER
>         bool "Android Low Memory Killer"
>         default N
>         ---help---
>           Register processes to be killed when memory is low
> 
> available in Staging drivers / Android?

It looks very similar to vanilla process selection to be killed, but it
gets into account only rss size before the oom parameters.

-- 
	Evgeniy Polyakov

Re: ANDROID low memory killer register (was Linux killed Kenny, bastard!)

[Evgeniy Polyakov]

On Tue, Jan 13, 2009 at 12:10:38PM +0000, Alan Cox (alan@lxorguk.ukuu.org.uk) wrote:
> I believe they are related in the sense that both of them are
> unneccessary and should not make the final kernel...
> 
> The /proc../oom_adj interface is again sufficient for this.

Besides the fact that it will unlikely to work the way you expect, since
oom adjustment is racy and may be too late, or not happend at all...

Alan, please try to apply this to practice and then you will suddenly
find that it would be really great to have a different method, which
will just work.

-- 
	Evgeniy Polyakov

Re: ANDROID low memory killer register (was Linux killed Kenny, bastard!)

[Alan Cox]

On Tue, 13 Jan 2009 15:23:56 +0300
Evgeniy Polyakov <zbr@ioremap.net> wrote:

> On Tue, Jan 13, 2009 at 12:10:38PM +0000, Alan Cox (alan@lxorguk.ukuu.org.uk) wrote:
> > I believe they are related in the sense that both of them are
> > unneccessary and should not make the final kernel...
> > 
> > The /proc../oom_adj interface is again sufficient for this.
> 
> Besides the fact that it will unlikely to work the way you expect, since
> oom adjustment is racy and may be too late, or not happend at all...

I would love to see a proposal that met such criteria. I don't think I
ever will and your random name based hack isn't it. The closest we have
right now is the container stuff.

Alan