Re: new memory = more swap?

From: Marc Schwartz (MSchwartz_at_MedAnalytics.com)
Date: 03/25/04

  • Next message: duncan brown: "Re: new memory = more swap?"
    To: For users of Fedora Core releases <fedora-list@redhat.com>
    Date: Thu, 25 Mar 2004 13:26:26 -0600
    
    

    On Thu, 2004-03-25 at 13:09, Matt Morgan wrote:
    > On 03/25/2004 01:50 PM, John Thompson wrote:
    >
    > >Craig Thomas wrote:
    > >
    > >>I have 256MB ram and a 502MB swap, and want to increase to 384MB
    > >>ram, [i know, i know it's an old machine]. I've read in the RH
    > >>manual and else where that double the amount of ram is "right". If
    > >>I want more swap but don't have any unpartitioned space left, what
    > >>are my options? (I do, however, have lots of free space on my
    > >>drive).
    > >>
    > >>
    > >The advice that swap=2(RAM) dates from the time when RAM was
    > expensive and few user machines had more than 64MB.
    > >
    > >These days most people have plenty of RAM and thus require less swap
    > space. I have 384MB RAM and a 256MB swap partition that is seldom
    > more than 25% used.
    > >
    > >
    > >
    > Is it still also true, though, that swap should at a minimum = RAM
    > (this
    > is knowledge that dates back to early versions of SCO, which was weird
    > anyway, I know)? If so, then having more swap than RAM may be
    > worthwhile
    > anyway, because you probably have plenty of disk space, and you might
    > add RAM later (and you won't have to repartition at that point if you
    > have extra swap).
    >
    > I realize it's also a lot easier to repartition these days, too. But I
    > still like to avoid it.
    >
    > Can you have too much swap? (Disk space issues aside). If I have 512
    > Mb
    > of RAM, and set up a 1Gb swap partition, did I make a mistake?

    I'm not sure that it would be a mistake. I suspect that the answer will
    be to a large extent specific to the use of the system such as what apps
    do you run, how many apps simultaneously, are you compiling apps, is
    this a server or workstation, what services do you run. Things like
    that.

    My older i8200 Dell laptop had 1GB of RAM and a 1GB swap. There were
    times when perhaps 30-40% of the swap space was in use.

    My new Dell i5150 laptop has 2GB of RAM, and I rarely ever see more than
    30-50 MB of swap in use and it is used largely in the same way as the
    i8200 was. As I am typing this, it is 0 MB in use.

    The guidelines are just that, guidelines. You need to have a better feel
    for your system's utilization and based upon that, adjust accordingly.
    Obviously, the more swapping that takes place, the slower the system
    response will be in general.

    HTH,

    Marc Schwartz

    -- 
    fedora-list mailing list
    fedora-list@redhat.com
    To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list
    

  • Next message: duncan brown: "Re: new memory = more swap?"

    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)
    • 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)
    • 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)
    • 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)