home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!sun-barr!sh.wide!wnoc-tyo-news!scslwide!wsgw!wsservra!daemon
- From: mosedale@genome.stanford.ed (Dan Mosedale)
- Newsgroups: fj.mail-lists.x-window
- Subject: Re: Why is xload such a memory pig?
- Message-ID: <1992Nov19.042459.28082@sm.sony.co.jp>
- Date: 19 Nov 92 04:24:59 GMT
- Sender: daemon@sm.sony.co.jp (The devil himself)
- Distribution: fj
- Organization: Yeast Genome Project, Stanford University, Stanford, CA
- Lines: 33
- Approved: michael@sm.sony.co.jp
-
- Date: 18 Nov 92 20:28:53 GMT
- Message-Id: <1992Nov18.202853.7625@leland.Stanford.EDU>
- Newsgroups: comp.sys.sun.apps,comp.windows.x,comp.unix.ultrix,comp.unix.admin
- References: <1992Nov11.223350.18090@leland.Stanford.EDU>
- Sender: xpert-request@expo.lcs.mit.edu
-
- >>>>> On Thu, 12 Nov 1992 19:02:54 GMT, jpiquer@dcc.uchile.cl (Jo Piquer) said:
- > In article <1992Nov12.023124.871@mnemosyne.cs.du.edu>, (Ian Hastie) writes:
-
- >> OK... I may be being stupid here, but as far as I understand it a
- >> programs RSS should always be smaller than it's SZ. At least surely
- >> it should not be bigger since this implies that the resident portion
- >> is bigger than the total size of the process. Is there really
- >> something wierd going on or is it just that ps work differently on
- >> Suns in this respect from the way it does on DEC's boxes??
- >
- > It's typical in Sun OS4.x, it is due to the shared libs, they are
- > counted for RSS but not for the SZ. xload includes a big part of the libX11
- > and libc.
-
- Aha. Now things are starting to make sense. This seems to make RSS a
- moderately useless statistic, however: if there are 6 xloads running,
- each with an RSS of about 1 meg, the total (actual) RSS of all the
- xloads will be much less than 6 megs.
-
- Does anyone know the rational behind this decision? More importantly,
- I guess, is there any way to access the "real" RSS of a process? Does
- sps do this?
-
- Thanks,
- Dan Mosedale
-
- mosedale@genome.stanford.edu
-