I have a PE 1600SC with a Perc 4/SC raid card running Windows 2003 standard.  The BIOS, Perc firmware, and drivers for the raid are all the latest versions.  Recently the server started getting sense errors type 3 on drive 0:4.  The problem is that I only had drives 0:0, 0:1, and 0:2 in the array running raid-5.  I had been wanting to add some additional storage anyway, so I bought 3 brand new drives for it that matched what was in the array (36 GB 10k 80 pin scsi).  I only was going to install 2 and keep the 3rd for an emergency drive since I have several servers that use that drive.  I inserted the 2 drives in 0:3 and 0:4.  Then rescaned the array adapter in the openmanage array software and it found the drives.  I added them to the array and it started to reconstruct the array.  The 0:4 drive then failed and I tried to rebuild it once thru the Perc 4 config software (CTRL-M at bootup) and also thru the openmanage array software.  Both times the drive would complete normally and things would be fine, but upon reboot the drive 0:4 would fail again.  I have tried all 3 drives and though it is possible for all brand new drives to be bad, I doubt it, espically since the computer thought that drive 0:4 was going bad before a drive was even in there.

I have searched for this and found one other person that was getting errors on a "ghost" drive.  They finally fixed it by removing all the drives and going into CTRL-M on bootup and deleting the container.  Then they shut the compuer off and reinstalled the drive and then told the Perc controller to reload the container info from the drives.

Anyone have suggestions on what to do to fix this problem?  If I try the above mentioned deletion of container and reloading the container info from the drives, does that delete all my data and make me reload from tape?