Hi Folks,

I have a server that is showing 4x4GB memories installed from the 'dmidecode' as below.

====================================================
dmidecode -t 17|grep "Size:"
Size: 4096 MB
Size: 4096 MB
Size: No Module Installed
Size: 4096 MB
Size: No Module Installed
Size: No Module Installed
Size: 4096 MB
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
Size: No Module Installed
====================================================

But only 4GB is detected.
In the dmesg, it also show 4Gb
====================================================

Linux version 2.6.18-128.4.1.el5PAE (mockbuild@builder16.centos.org) (gcc v
ersion 4.1.2 20080704 (Red Hat 4.1.2-44)) #1 SMP Tue Aug 4 20:58:34 EDT 200
9
BIOS-provided physical RAM map:
BIOS-e820: 0000000000000000 - 000000000009c800 (usable)
BIOS-e820: 000000000009c800 - 00000000000a0000 (reserved)
BIOS-e820: 00000000000e4000 - 0000000000100000 (reserved)
BIOS-e820: 0000000000100000 - 00000000bf7c0000 (usable)
BIOS-e820: 00000000bf7c0000 - 00000000bf7cb000 (ACPI data)
BIOS-e820: 00000000bf7cb000 - 00000000bf7cc000 (ACPI NVS)
BIOS-e820: 00000000bf7cc000 - 00000000c0000000 (reserved)
BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved)
BIOS-e820: 00000000fec00000 - 00000000fed00000 (reserved)
BIOS-e820: 00000000fee00000 - 00000000fee01000 (reserved)
BIOS-e820: 00000000ffa80000 - 0000000100000000 (reserved)
BIOS-e820: 0000000100000000 - 0000000140000000 (usable)
4224MB HIGHMEM available.
896MB LOWMEM available.
found SMP MP-table at 000f8440
Memory for crash kernel (0x0 to 0x0) notwithin permissible range
disabling kdump
.
.
.
Memory: 4135296k/5242880k available (2120k kernel code, 49224k reserved, 88
1k data, 228k init, 3268352k highmem)

====================================================

Can anyone advise me on this?
Thanks

If it's a 32-bit box, 4G is all it will use. If a physical address extension (PAE) kernel is available for your distribution, install that. Otherwise you can rebuild your kernel with PAE turned on. There's a (Ubuntu) discussion about this at http://www.cyberciti.biz/faq/ubuntu-...tion-solution/.

Hope this helps some.

[EDIT]
Well, duh! Missed already PAE (old, blind in one eye and can't see out the other one).

johnssfine has the right ideas -- BIOS, bad RAM strip (not seated RAM strip?), etc.

Hate gettin' old...
[/EDIT]

0 members found this post helpful.

The original post clearly indicates that it is already PAE.

The memory map (if you know how to read it) indicates the problem is not in the Linux software. A 64 bit kernel would have no more access to this ram than the 32 bit PAE kernel has.

What information do you have on the motherboard? (From purchase, or dmidecode or where ever).

My first estimate is that this is a motherboard problem. But from there I'm not expert enough to diagnose further. Maybe some of the ram is defective. Maybe you need a newer BIOS version. Maybe you just need to change a BIOS setting. I would barely know how to distinguish the various motherboard/ram/BIOS possibilities if I were working the issue on site. So I won't be much help on that from a distance. But I am confident it is a motherboard, ram, or BIOS issue and not a Linux issue.

BTW, 16GB is pushing the limit for effective use of PAE. Depending on the mix of applications you run, it might by OK or it might exhaust the 1GB limit on kernel virtual memory.

When/if you solve the hardware/BIOS problem, I think you would be better off with a 64 bit kernel. A 64 bit kernel will not help solve the problem you posted about, and after that problem is fixed a 32 bit PAE kernel can use 16GB. But a 64 bit kernel might be significantly better at using 16GB, even if all your applications are 32 bit.

Hi,

Here is the system info. Thanks.
=====================================
# dmidecode 2.7
SMBIOS 2.5 present.
92 structures occupying 3247 bytes.
Table at 0xBF7D9000.

Handle 0x0000, DMI type 0, 24 bytes.
BIOS Information
Vendor: FUJITSU // Phoenix Technologies Ltd.
Version: 6.00 Rev. 1.10.2786
Release Date: 06/23/2009
Address: 0xE4890
Runtime Size: 112496 bytes
ROM Size: 2048 kB
Characteristics:
PCI is supported
PNP is supported
BIOS is upgradeable
BIOS shadowing is allowed
ESCD support is available
Boot from CD is supported
Selectable boot is supported
EDD is supported
Print screen service is supported (int 5h)
8042 keyboard services are supported (int 9h)
Serial services are supported (int 14h)
Printer services are supported (int 17h)
CGA/mono video services are supported (int 10h)
ACPI is supported
USB legacy is supported
BIOS boot specification is supported
Targeted content distribution is supported
BIOS Revision: 1.10

Handle 0x0001, DMI type 1, 27 bytes.
System Information
Manufacturer: FUJITSU
Product Name: PRIMERGY RX200 S5
Version: GS02
Serial Number: ****
UUID: 00000000-0000-0000-0000-000AE48B8890
Wake-up Type: Power Switch
SKU Number: ABN:K1272-V101-6
Family: SERVER


===================================

PAE is installed

===================================
Handle 0x0004, DMI type 4, 40 bytes.
Processor Information
Socket Designation: CPU 1
Type: Central Processor
Family: Xeon
Manufacturer: Intel
ID: A5 06 01 00 FF FB EB BF
Signature: Type 0, Family 6, Model 26, Stepping 5
Flags:
FPU (Floating-point unit on-chip)
VME (Virtual mode extension)
DE (Debugging extension)
PSE (Page size extension)
TSC (Time stamp counter)
MSR (Model specific registers)
PAE (Physical address extension)
==========================


Thanks again, johnsfine and tronayne

Maybe the full dmidecode -t 17 would shed some light (rather than just the size lines that you posted).

I looked at section 6 of your manual:
http://manuals.ts.fujitsu.com/file/8...00s5-og-en.pdf
That has various rules for populating memory that should be compared against whatever dmidecode tells you.

I have compared this server with another that has the same spec, even the positions of memory module installed are the same.
That server detects 16GB instead of 4GB.
This is really weird.

more details of dmidecode -t 17:


Handle 0x0029, DMI type 17, 27 bytes.
Memory Device
Array Handle: 0x0028
Error Information Handle: Not Provided
Total Width: 72 bits
Data Width: 64 bits
Size: 4096 MB
Form Factor: DIMM
Set: None
Locator: DIMM-1A
Bank Locator: Not Specified
Type: Other
Type Detail: Synchronous
Speed: 800 MHz (1.2 ns)
Manufacturer: Not Specified
Serial Number: Not Specified
Asset Tag: Not Specified
Part Number: Not Specified

Handle 0x002A, DMI type 17, 27 bytes.
Memory Device
Array Handle: 0x0028
Error Information Handle: Not Provided
Total Width: 72 bits
Data Width: 64 bits
Size: 4096 MB
Form Factor: DIMM
Set: None
Locator: DIMM-2A
Bank Locator: Not Specified
Type: Other
Type Detail: Synchronous
Speed: 800 MHz (1.2 ns)
Manufacturer: Not Specified
Serial Number: Not Specified
Asset Tag: Not Specified
Part Number: Not Specified
Handle 0x002C, DMI type 17, 27 bytes.
Memory Device
Array Handle: 0x0028
Error Information Handle: Not Provided
Total Width: 72 bits
Data Width: 64 bits
Size: 4096 MB
Form Factor: DIMM
Set: None
Locator: DIMM-1B
Bank Locator: Not Specified
Type: Other
Type Detail: Synchronous
Speed: 800 MHz (1.2 ns)
Manufacturer: Not Specified
Serial Number: Not Specified
Asset Tag: Not Specified
Part Number: Not Specified
Handle 0x002F, DMI type 17, 27 bytes.
Memory Device
Array Handle: 0x0028
Error Information Handle: Not Provided
Total Width: 72 bits
Data Width: 64 bits
Size: 4096 MB
Form Factor: DIMM
Set: None
Locator: DIMM-1C
Bank Locator: Not Specified
Type: Other
Type Detail: Synchronous
Speed: 800 MHz (1.2 ns)
Manufacturer: Not Specified
Serial Number: Not Specified
Asset Tag: Not Specified
Part Number: Not Specified