[RndTbl] Debian doesn't see all RAM in a VM

Adam Thompson athompso at muug.mb.ca
Mon Feb 1 23:17:10 CST 2010

On 2010-Feb-01 23:08, Sean Cody wrote:
> Can you post a dmesg?
> That would help lots.
> I'm assuming the machine's POST mem test gets to 2048 right?
> 4GB has issues on some architectures (memory hole due to address space layout) but 2GB sounds odd and I would need more detail to even guess.

Agreeing completely with Sean here - and 900Mb is just plain a very 
strange number that makes even less sense than 1024Mb.

In addition to dmesg(1) output, the dmidecode(8) tool can tell you what 
the SMBIOS has emulated in this particular VM; limit the output to 
memory modules by select DMI Type 17, i.e. "dmidecode -t17"
Obviously the output doesn't relate to real physical sockets, but the 
SMBIOS (aka the DMI table) should still report consistent information. 
For that matter, I wouldn't mind seeing the full dmidecode(8) output 
from an ESXi VM - never occurred to me to look and I don't have access 
to an ESXi system anymore.


More information about the Roundtable mailing list