Re: Top - Mem: 119212k used

From: Robert Heller (heller_at_deepsoft.com)
Date: 07/29/03 

Date: Tue, 29 Jul 2003 18:39:28 +0000

  Alan Connor <xxxxxx@xxxx.xxx>,
  In a message on Tue, 29 Jul 2003 17:29:54 GMT, wrote :

AC> On Tue, 29 Jul 2003 16:24:09 GMT, Ed Murphy <emurphy42@socal.rr.com> wrote:
AC> >
AC> >
AC> >> Freed from WHAT? It's not being used for anything anyway.
AC> >
AC> > I think they meant "freed from being used as cache" - but I think they
AC> > may have also misunderstood "60+megs" as "60+megs including cache". You
AC> > meant "60+megs after excluding cache", right?
AC> >
AC> > Here's some data from your original /dev/vcs2 examination:
AC> >
AC> > 126528K total
AC> > 7316K free
AC> > 18932K buffers
AC> > 45920K cached
AC> >
AC> > This indicates 54360K in use, after excluding buffers and cache. Have
AC> > you itemized this usage by process?
AC> >
AC>
AC> Been working on it, Ed.
AC>
AC> I removed all the gettys but 2 from /etc/inittab, and when running from a tty
AC> with screen, as I am at present, (X not up) I am using about 12 Megs of RAM,
AC>
AC> When I bring up the xserver, open an xterm and use it to launch a graphical
AC> browser (skipstone: REALLY need something smaller that's still graphical!)
AC> mem usage goes up to 37 megs! (16 for the xserver alone. i'm using flwm, which
AC> is the smallest window manager I could find.

On my system:

sauron.deepsoft.com% free
             total used free shared buffers
cached
Mem: 126744 123080 3664 0 2676 38608
-/+ buffers/cache: 81796 44948
Swap: 530136 68920 461216

I use fvwm2 as my window manager. I use real xterms
(/usr/bin/X11/xterm, not the 'fancy' xterms), and Mozilla. I have both
Postgres AND Apache running. I *don't* use any sort of graphical file
manager and use MicroEmacs 3.10 for editing.

AC>
AC>
AC> Alan
AC>
AC>
AC>
AC> --
AC> For Linux/Bash users: Eliminate spam with
AC> the Mailbox-Sentry-Program. See the thread
AC> MSP on comp.mail.misc for the scripts and docs.
AC>
AC>

                                     \/
Robert Heller ||InterNet: heller@cs.umass.edu
http://vis-www.cs.umass.edu/~heller || heller@deepsoft.com
http://www.deepsoft.com /\FidoNet: 1:321/153

          

Relevant Pages

  • Re: Informix FAQ
    ... BTR is an attempt to quantify how much or little churning of the buffer ... That is deivided by the number of buffers in the cache and the ... The UBTR was a largely unsuccessful attempt to eliminate unused buffers from ... See you at the 2010 IIUG Informix Conference ...
    (comp.databases.informix)
  • RE: Informix FAQ
    ... portion of the cache was replaced. ... Here's my reasons for using pagreads and bufwrits. ... The other side of the BTR are the pagreads. ... Your pagwrits indicate that you wrote 2,620,000 buffers per hour to disk. ...
    (comp.databases.informix)
  • Re: does a CPU with hit-under-miss always imply that a microarchitecture has refill buffers?
    ... counting the number of refill buffers. ... buffers than the number of cache misses tolerated. ... Is this a CPU microarchitectural axiom? ... the small number of "refill buffers" is a bottleneck. ...
    (comp.arch)
  • Re: Increased Buffers due to patch 56e49d (vmscan: evict use-once pages first), but why exactly?
    ... All is about the increased amount of "Buffers" accounted as active while ... But I didn't get how this prefers buffers compared to cache pages (I ... This indeed sounds like the kind of workload that would only ... it might makes stream I/O benchmark score a bit because such workload ...
    (Linux-Kernel)
  • Re: Increased Buffers due to patch 56e49d (vmscan: evict use-once pages first), but why exactly?
    ... All is about the increased amount of "Buffers" accounted as active while ... But I didn't get how this prefers buffers compared to cache pages (I ... This indeed sounds like the kind of workload that would only ... filesystem metadata all the time. ...
    (Linux-Kernel)