OT: free cmd is lying to me

[Holger Rauch <Holger.Rauch@heitec.de>]

Hi!

First of all, I admit that this is slightly OT, since I believe this to be
not strictly related to Debian, but rather a general mainboard/kernel/mem
config issue.

I got a system equipped with a Supermicro dual processor mainboard, one 
PIII 1GHz CPU (the other processor slot is currently unused) and 2048 MB
main memory, which the kernel recognizes correctly upon system boot. In
addition, I got two swap partitions configured, one on an ordinary
partition, the other one on a LV, giving me 4 GB swap space
total. I am running a 2.2.19 SMP kernel (including the driver for the
ICP Vortex RAID controller) and Oracle 8.1.7 on that
machine. Oracle's db_block_buffers are configured so that Oracle uses 600
MB for the DB. Apart from Oracle, only the following daemons are running
on the system:

- cron
- syslogd
- xdm

That said, the free cmd should report that approx. 1.4 GB of main memory
should still be available. However, what I get is:

             total       used       free     shared    buffers     cached
Mem:       2096664    2093416       3248          0      87172    1783556
-/+ buffers/cache:     222688    1873976
Swap:      4194272          0    4194272

In order to find out whether this is right, I wrote a small test
program that continuously does a malloc() for 10 MB every 5 secs until the
system runs out of memory. Strangely enough, it didn't. I did a "free -s
5" on the system to see whether the system is going to use swap space when
running out of main memory. Not the case. The column for used swap space
*always* showed 0. Also, the column for used mem didn't change
significantly (just increased by 4 to 5 MB.) Now my questions are:

1. What could be the cause that size of used mem doesn't increase
accordingly when I malloc() 10 MB?

2. Why does "free" leave the impression on me that no swap space is used?

3. Are there any known problems with memory detection on Supermicro
mainboards?

4. Are there any processes within Oracle 8.1.7 that are known to have mem
leaks? If so, which of them are the culprits?

Thanks for taking your time! Any help will be greatly appreciated! 

Kind regards,

	Holger

P. S.: I've never seen such an effect before and I'm not familiar with
Supermicro mainboards. Up to now, I've mostly used Asus and Gigabyte
mainboards.