topic Re: ML570 G2 ECC Memory Correctable Errors detected in ProLiant Servers (ML,DL,SL)

ML570 G2 ECC Memory Correctable Errors detected

masterking — Mon, 10 Jan 2011 19:16:45 GMT

Hello,

I've a ML570 G2 with OS Linux Suse.
HP Agent send me this alert:

Trap-ID=6056
ECC Memory Correctable Errors detected.

How can I find the bench faulty ram from command line ?

Regards.

Re: ML570 G2 ECC Memory Correctable Errors detected

Michael A. McKenney — Mon, 10 Jan 2011 19:21:32 GMT

I would reboot and boot the smart start and run diagnostics. Other way is memtest86 bootable CD and test one stick at a time.

Re: ML570 G2 ECC Memory Correctable Errors detected

masterking — Tue, 11 Jan 2011 04:56:27 GMT

Thank you Michael,

what are the tools to do an analysis before restarting the server?

Thanks again

Re: ML570 G2 ECC Memory Correctable Errors detected

Johan Guldmyr — Tue, 11 Jan 2011 06:42:52 GMT

You want to test the memory "offline" - by booting on a CD, because then as little as possible memory is used by the OS.

If you want to run tests from inside Suse you can try the HP Online Diagnostics.

See
http://h20000.www2.hp.com/bizsupport/TechSupport/DriverDownload.jsp?prodNameId=3285471〈=en&cc=us&taskId=135&prodTypeId=15351&prodSeriesId=316549

and select your Suse OS version and "HP Insight Diagnostics Online Edition for Linux".

Re: ML570 G2 ECC Memory Correctable Errors detected

Michael A. McKenney — Wed, 12 Jan 2011 00:50:49 GMT

http://www.memtest86.com/download.html

Download the ISO and burn it to a CD. Install one stick at a time and boot the CD. Do complete tests on each stick. Any stick fails, replace. Its the best memory test I have seen in 25+ years. I use it.