server refuses to accept a ram-reduction 1

[netcomander]

hi all,

I'm running a linux server under RedHat 7.0.
These day's I gave it more RAM (2 X 512MB).

But there were several problems wich came
along with this change of state, and thus I
wanted to verify, whether this could have somethin
to do with a RAM-flaw.

After removing 512MB, the servers attempt of
booting ends up in a Kernel-panic. Besides, according
to what the screen shows up, the server is obviously
assuming, that it's still got (1024 MB).

Does anybody know, where the server obtains it
information of how much RAM is available?

would be nice when getting a hint

greetings markus

 

[ifincham]

Hi,

To see what memory the kernel thinks you have do :

cat /proc/meminfo

Sometimes it does get it wrong - usually the other way round actually - however you can feed the kernel with a memory value at boot time either by passing a argument to the boot image name or, more permanently, by adding an 'append' to /etc/lilo.conf which will do that automatically . For example :

append="mem=512M"

that can be 'global' or within an 'image' block like this :

image=/boot/vmlinuz-2.4.17
label=linux
root=/dev/hda7
append="mem=512M"
read-only

The other way is to add at the boot: prompt - for example (control-X from graphical lilo screen to get the boot: prompt) -


boot: linux mem=512M

Hope this helps