I have a few DL 380 G5 servers running esx 4.1 that I'm getting ready to lock away at a remote site 300 miles away. As I'm finishing up the install, I have a memory light flashing on the front panel of the server. Ok, I have a bad DIMM or whatever, but it got me thinking how would I know about this from afar? First, I went in to vcenter and looked at the hardware status of the server, and and the offending DIMM just reads the same as all the others, reporting what it is, so that's not going to work. I have the iLo configured so I went in to that and looked at the memory tab in system health, and it just reports all the DIMMS in there and their specs. That's not going to help. I have Insight Manager installed, but my thought is that if the iLo isn't seeing a problem then Insight Manager isn't going to see it either.
Does anyone have any experience with this? Why is there a disconnect between what the front panel is reporting and what iLo says? I don't get that.
Also, reseating the DIMM makes the light go away for a few minutes, then comes back, so it's probably bad. I'd just like to be able to find this out remotely.
Thanks!
the only way you are going to get this in depth monitoring is to install the HP Insight Manager Agents, from there redirect the SNMP traps to SIM. vCenter does not have the proper CIM modules for this type of hardware monitoring.
Thanks for the reply. Do have any idea why there's a disconnect between what the front panel is saying and what iLo is saying?
Thanks!
To be honest, I have seen conflicting information from both the front panel LED and iLO2. You may try a few things. I would first ensure all your firmware is current, then maybe boot up into the SmartStart Diagnostics CD and run some hardware diagnostics. Also, if there is a bad DIMM, usually a reboot will catch it, and your POST screen will let you know. I would compare the POST to what the front panel LED says, and then look at replacing the failed component.
