Re: Swap usage always goes up - never down

From: Chris (ithinkiam_at_gmail.com)
Date: 08/19/05 

Date: Fri, 19 Aug 2005 15:25:31 +0100

Michael Heiming wrote:
> In comp.os.linux.misc Chris <ithinkiam@gmail.com>:
>
>>Michael Heiming wrote:
>>
>>>In comp.os.linux.misc Dan Espen <daneNO@spam.mk.telcordia.com>:
>>>
>>>
>>>>Michael Heiming <michael+USENET@www.heiming.de> writes:
>>>
>>>
>>>>>In comp.os.linux.misc Dan Espen <daneNO@spam.mk.telcordia.com>:
>>>>>
>>>>>
>>>>>>"Peter T. Breuer" <ptb@oboe.it.uc3m.es> writes:
>>>>>>
>>>>>>
>>>>>>>Dan Espen <daneNO@spam.mk.telcordia.com> wrote:
>>>>>>>
>>>>>>>
>>>>>>>>"Peter T. Breuer" <ptb@oboe.it.uc3m.es> writes:
>>>>>>>
>>>>>>>>>Chris <ithinkiam@gmail.com> wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>Peter T. Breuer wrote:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>>Chris <ithinkiam@gmail.com> wrote:
>>>>>
[snip]
>
>>But, if I'm querying whether I'm running out of swap occasionally surely
>>settting overcommit_memory to positive would result in more problems,
>>not less?
>
>
> It depends, you'd need to profile the memory usage of working
> apps, easiest would be to just try it out and monitor the system
> closely (sar/mrtg/etc). Then decide if it settings are fine,
> there are a bunch of other settings allowing to tune vm
> behavior, the question is usually if it's worth all the work or
> if just adding RAM, which is cheap enough or even more swap is
> enough to solve the problem.

I'm pretty sure I'm not running out of ram as I've always got
buffers/cache available in RAM, even when I'm running my 'big' jobs.

Currently I've only got thunderbird running and this is what free shows:

            total used free shared buffers cached
Mem: 513776 436040 77736 0 80732 245196
-/+ buffers/cache: 110112 403664
Swap: 1028120 73584 954536

I've even killed things like amarok, suse plugger, OO quick starter, but
I'm still using >70Mb of swap. I realise these are apps that are no
longer active, but I can't think what else to kill. All that's left in
ps of any significance is kdeinit stuff, nscd and X. Could there be a
leak somewhere?

> Can't look into your system, so it's not possible to tell what
> really needs to be done. In the mean time, I'd just try it out
> which should show you quite fast. What's the problem with just
> doing it?
>

Lack of root access! :-( And the sysadmin is very workshy... If the
system is running then he can't see a problem.

Relevant Pages

  • Re: excessive swap-in time
    ... Trying to do it all with RAM is just inefficient. ... I find that does not work well when the swapping is caused by excessive I/O ... Or you could have 5 TB of swap space. ...
    (comp.os.linux.development.system)
  • SLUB 0:1 SLAB (OOM during massive parallel kernel builds)
    ... 2G of RAM, 1G of swap partition. ... DMA per-cpu: ... kill process 25286 score 188662 or a child ...
    (Linux-Kernel)
  • fork failed cannot allocate memory
    ... I have a Dell 2650, dual Xeon box, 2GB RAM, 6GB swap with PERC Hardware ... feeding the perl processes). ... Every week or two the box will stop allowing new netcool client ...
    (RedHat)
  • 2.6.26-rc5-mm2: OOM with 1G free swap
    ... OOM condition happened with 1G free swap. ... 4G RAM, 1G swap partition, normally LTP survives during much, much higher ... Call Trace: ... 675611 total pagecache pages ...
    (Linux-Kernel)
  • Re: 2.6.0-test9 - poor swap performance on low end machines
    ... >> Well I was considering adding the swap pressure to this algorithm but I ... It wont be as aggressive as setting the swappiness ... > The test compile started in a similar way to the compile when using your ... detriment of other tasks that are on the runqueue and still need ram. ...
    (Linux-Kernel)