ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

ML370 G4 RAID expansion

Mike Asp
Occasional Visitor

ML370 G4 RAID expansion

I added a 4th drive to 3 72G U320 15k drives in a RAID5. The expansion process started normally and ran for about 14 hours ~ 68% complete until the server crashed apparently due to an array timeout. POST indicated that several drives had been "removed" and the array was lost, hit F2 to accept the "data loss" and continue. The server booted and ran a lengthy chkdsk.

I noticed that within a few minutes after booting, all the error, status and activity lights on all 6 drives (incl 2 in a mirror) would randomly and erratically flash and turn off. esp during high activity. Sometimes the flashing would stop, but the status lights were then off. only activity lights working. Eventually the server would crash and report that drives had been removed. - This server never crashes.

I replaced the controller with a spare - same result and removed the new drive - same result.

I did all firmware/Driver/software updates possible.

HP advised to recreate the array and restore the data from backup. No joy, still same symptoms. Then they advised to run diagnostics. During diagnostics, the array spazzed out. I sent the diag reports to HP and they advised that the new drive model/firmware was incompatible with the existing drives. (say what?) Sending new one. They want me to reset CMOS and recreate the array again. I have not received it yet.

After leaving it for the weekend, expecting it would crash the server remained up for 3 solid days now and is now acting completely normal. Normal lights, everything fine.

Anyone see this before? Ideas?