Proliant 1600 Parity error problem.

[Todd Lilley]

We have a Proliant 1600 with single 600 PIII and 512 MB ram. This was a production server for several years and now has been replaced by a new Proliant.
We want to use this 1600 as a Windows SUS server for our workstation updates as well as service pack storage. I rebuilt the system and it ran for several weeks.
After about 2 or 3 weeks of up-time, we started getting parity errors with the ram. We had installed 1 256MB and 2 128MB DIMMS all PC100. I ran memtest86 several times and found 1 dimm was bad. the other two passed. removed the bad 256MB DIMM and it ran for a day or two until still another parity error. ran memtest again and removed another DIMM which tested bad. We're now down to 128MB of ram. Our server was up for less than a day before the last DIMM caused another parity error.
I just can't believe that 3 DIMMS would go bad this fast. I'm thinking of a system board problem or something else of this nature. Does anyone have any ideas?

[JohnWRuffo · Accepted solution]

Todd:

With so many DIMM's going bad, yes, I would agree there may be a problem with the CPU Cage.
First thing to do is pull the cage out and thuroughly clean it of dust and such. Make sure you get the RAM sockets blown out well.

Then re-try the box with the RAM back in it.

If you still get errors, you may need to replace the cage:
http://partsurfer.hp.com/cgi-bin/spi/showphoto?partnumber=313622-001

Let me know what you find? -john

[Todd Lilley]

John;
Thanks for the reply. I have my server dusted out, I re-seated the processor and ram, and it's now booted. We'll see how long it stays up this time.

Thanks,

Todd

[Josh Harper]

On this server was you running 2003 because i am recieving the same error and have switched memory too if the cleaning seemed to fix it please let me know

[Todd Lilley]

Hi Josh;

No, the cleaning didn't fix my problem. We are running win2k server. We are in the process of purchasing a new/reconditioned system board for our 1600.

Todd