DL385 G1 ESX 3.01 host crashes with ASR, no errors logged

[dcs_1]

I have a production DL385 running VMWare ESX 3.0.1 (16GB RAM, 12 guests, 2x Opteron 280, local storage) which just crashed the other night. The only error in the integrated management log is "ASR Detected by System ROM".

I presume there's a hardware problem, since I have 9 other identical DL385's with ESX 3.0.1, same host OS configuration and equal or greater loads. When I've looked at the VMWare forums, they always point back to hardware - test your RAM, etc.

Are the SmartStart CD tests the best way to dig deeper and find underlying problems? Or are there better tools in the BIOS/ROM?

[David Claypool]

The fact that an ASR was logged is the error. This occurs when the system has become unresponsive and has not come back after 2 or more queries by the ROM watchdog timer.

[dcs_1]

Thanks for the quick reply. Since ESX doesn't show any errors before the crash, I'm leaning toward hardware diagnostics. Are the SmartStart, offline diagnostics the most comprehensive?

[Michael Richter]

Yes it is. Please be sure to use the latest. I think the Ram Test is much more efficient. You might also try 'memtest86'

cheers
mike