sportmode wrote in post #12703171
I don't know if Drobo provides this, but see if you can get the SMART info on each of the drives -- this will tell you if each of the drives are actually clean or if they've encountered uncorrectable errors. That'll give you a clue as far as if the errors were drive-level or higher up.
If Drobo do not support SMART info and their support are not able to figure out if the disks are physically ok or not, then Drobo should be dropped from the market. No, I really can't see them not having built-in support for SMART supervision.
Then you get to the filesystem level. I assume you're using NTFS. There's a possibility that your disks are all fine, your Drobo RAID acted appropriately, but your filesystem crapped out. Usually I've seen cases where the memory on the PC has started to get flakey, and that in itself could've started messing with your data. Keep in mind that since the Drobo is a direct-attached device, it will be using the memory on your PC to read and write data to it directly.
This is one of the major reasons for not seeing a RAID as a backup replacement. When the RAID controller gets the "wrong" write command, it will perform that command. Being a RAID, the write will affect all disks in the volume. The end result will be similar to a bad write command to a single disk, or a simple disk mirror.
But first step with data recovery is to do image backup of all your disks -- you have to remember that some data recovery process can be destructive and can actually eliminate any possibility of recovery. Unfortunately, unless you have some spare drives with large enough capacity you can image to as well as a PC setup to do this, there's certainly some cost and time involved with it.
Taking a raw copy of the partitions and performing the recovery on this copy is most definitely the #1 step when a disk/file system is in dire need of repair. This is a reason why huge partitions are not always a good idea, even if it is easier to work with a continuous partition than to have to figure out which partition that contains a specific directory or how much more that fits in the remaining space of a partition.
1) A huge partition that fails may bring lots of files with it into the grave.
2) Having a need to take a copy of a huge multi-disk partition can be really nasty if the involved computer don't have interfaces for enough disks to span the partition. And it takes a RAID controller or a volume manager to be able to concatenate multiple backup disks into a big enough surface to swallow the original partition.