Memory usage

Steve

unread,

Nov 13, 2002, 7:16:16 PM11/13/02

to

Is there anyway to check how much physical memory is being used and any
point in time?

Thanks,

steve

Stuart J. Browne

unread,

Nov 13, 2002, 7:47:26 PM11/13/02

to

"Steve" <mc...@qwest.net> wrote in message
news:sXBA9.185$f43....@news.uswest.net...

> Is there anyway to check how much physical memory is being used and any
> point in time?

sar -r

See 'man sar' for more details. If you've not previously enabled sar,
you'll have to execute "/usr/lib/sa/sar_enable -y" to start logging.

bkx

John DuBois

unread,

Nov 14, 2002, 8:42:21 PM11/14/02

to

In article <aqurum$m8u$1...@perki.connect.com.au>,

Or if you want to know how much is being used "right now":

echo o -d freemem|crash

Subtract the number this gives you from the amount of physical memory on the
system.

Also see: ftp://ftp.armory.com/pub/admin/memuse

John
--
John DuBois spc...@armory.com KC6QKZ/AE http://www.armory.com/~spcecdt/

Steve

unread,

Nov 15, 2002, 12:48:34 PM11/15/02

to

Thank you for the advice. I ran the echo command while the server was
behaving normally with the following results f0199988 0000016920, then when
the maching is acting up f0199988 0000000060. The machine has 96MB of ram
memory.

Steve

"John DuBois" <spc...@deeptht.armory.com> wrote in message
news:3dd450fc$0$22477$8ee...@newsreader.tycho.net...

John DuBois

unread,

Nov 15, 2002, 9:38:33 PM11/15/02

to

In article <4saB9.58$zZ5....@news.uswest.net>,

Steve <ebuzzm...@yahoo.com> wrote:
>Thank you for the advice. I ran the echo command while the server was
>behaving normally with the following results f0199988 0000016920, then when
>the maching is acting up f0199988 0000000060. The machine has 96MB of ram
>memory.

Right. I should have mentioned that freemem gives available real memory in 4K
pages, so the first output you got indicates 16920*4 = 67680KB available.

Steve

unread,

Nov 18, 2002, 11:50:58 AM11/18/02

to

The server went crazy again late friday so I ran ps -elf with the following
result:
vfsd --profile /usr/vision/vfsprofile and the SZ = 313032.

Is this also in 4K pages. The user associated with this process is a
windows XP machine.

Thank you,

steve

"John DuBois" <spc...@deeptht.armory.com> wrote in message

news:3dd5afa9$0$79559$8ee...@newsreader.tycho.net...

John DuBois

unread,

Nov 20, 2002, 7:56:24 PM11/20/02

to

In article <gU8C9.20$td2....@news.uswest.net>,

Steve <ebuzzm...@yahoo.com> wrote:
>The server went crazy again late friday so I ran ps -elf with the following
>result:
>vfsd --profile /usr/vision/vfsprofile and the SZ = 313032.
>
>Is this also in 4K pages.

No, it's in units of 1 KB. So, this process all by itself is using several
times the amount of real memory you have available, and is presumably making
the system swap madly.

Bela Lubkin

unread,

Nov 21, 2002, 4:34:53 AM11/21/02

to sco...@xenitec.on.ca

Steve wrote:

> The server went crazy again late friday so I ran ps -elf with the following
> result:
> vfsd --profile /usr/vision/vfsprofile and the SZ = 313032.
>

> Is this also in 4K pages. The user associated with this process is a
> windows XP machine.

`ps -l` size output is in 1K units. So that process was taking over
300MB, and was probably still growing as fast as possible (or dumping
core due to having grown too large).

Make sure you have the newest version of VisionFS, and if it's still
flaking out, talk to Tarantella about it.

>Bela<