I have an older HP Proliant ML110 G3 server running Windows 2003 Server that has exhibited a problem I have never encountered before and am wondering if anyone has seen this before.
The server was setup with a RAID 1 array on the embedded SATA RAID controller. It has been running fine for years but last week I logged into the server and noticed a bunch of Add New Hardware Wizards had popped up. The system was installing IDE controllers. I found that instead of 1 boot driver I know had both a C: Drive and an E: drive. Both with the same data. Clearly the RAID array had broken and the two drives where now recognized by the system a separate drives.
I rebooted the system to get into the Embedded RAID controller setup but found it was not there. In fact, I had to go into the system BIOS and re-enable the RAID option on the IDE Configuration. So, what happened was that something caused the system BIOS to change the IDE Configuration option from RAID to IDE.
I feel extremely confident this was due to some kind of human problem so the only thing I can think of is that the CMOS battery failed which forced the BIOS to reset everything to defaults.
I rebuilt the RAID array and got the system stable again. Then, I upgraded the system BIOS. After that was done I replaced the battery expecting the IDE Configuration setting to switch back to IDE from RAID.
Here is where it gets even more interesting. After putting in the new battery, the setting DID NOT change. That is what puzzles me. Either the upgraded BIOS somehow made that RAID setting "sticky", which I didn't think could happen, or something other than the CMOS battery failing caused the initial change in settings.
One other note, the old battery checked out at 2.8 volts and not 3.0 volts.
Any other ideas on what might have happened here? This one has me stumped.