[tato386]

I have an IBM rackmount server I suspect of having some kind of hardware problem. I have used Win2003 and also Win2008 with just a few apps and in both cases I get random bluescreens every couple of weeks. The BSOD messages are always different and always memory related. Some examples are: bad_pool_caller, page_fault_in_nonpaged_area and driver_left_locked_page_in_process.

I am trying to use UBCD to see if I can detect memory errors. However I have two issues:

1) If I use the Memtest from the main menu (DOS version) Memtest hangs after a couple of minutes
2) If I use Memtest from the UBC Windows I can only test a few hundred MB. In this mode only 2gb are recognized. (The server had 8GB of RAM)

What other options do I have here?

Thanks,
Diego

This post has been edited by tato386: 14 November 2009 - 04:27 PM

[rdsok]

This is another issue that has many causes and I've only listed a few here.

The most crictical possible cause you need to check for at first is if the system's CPU and other cooling fans are working properly or not. This is very important to check since if this has happened it could lead to hardware failure. Closely check that all of the fans in the system are clean and working properly. If they are dirty, clean them out with can of compressed air, if they have failed and quit working replace them before you turn on your computer again. A failed or even dirty fan can result in other components failing completely. Remember that cleaning your computer regularly is part of its regular maintenance in order to keep it in working order.

Also open the system case up and look for leaking or bulging capacitors.

You may also want to test will all non-essential components removed from the system and then add them back in one at a time to see if you can isolate the issue to one of these.

[tato386]

I agree totally with your points. However the server is new, very clean and has been doing this out of the box since we got it about 6 months ago. That is why I am looking into hardware issue.

Thanks,
Diego

[mbarnes]

Hi tao386

memtest86 and memtest86+ are different products from different teams

maybe you can find one which will recognises more than 2GB of ram on your server

if not you will have to remove most of the ram chips and test 2GB at a time

Does the ram have error correction ?
On some servers (HP ML-370) the error correcting ram is tested at startup and a front panel led comes on if too much bad ram is found

Also check for loose connections such as PSU to motherboard cables etc

regards
Mike Barnes

[rdsok]

tato386, on 15 November 2009 - 10:00 AM, said:

I agree totally with your points. However the server is new, very clean and has been doing this out of the box since we got it about 6 months ago. That is why I am looking into hardware issue.

Thanks,
Diego

I guess you didn't notice or perhaps didn't understand... the things I suggested to check ARE hardware

[tato386]

I guess I forgot the part about the capacitors which I will check. However all your other points have been addressed. All fans are clean, running well and at the proper RPM. The system has eight CPU fans and two fans for PSU. All are OK. Server room is kept at 68 degrees to boot.

I downloaded Memtest86+ V4.00 and this guys sees all memory and does not hang. I will run this overnight to see what happens. Hoping that it finds something.

Thanks,