Whats the point of raid if raid card failure resulted in complete loss of data? I came across an article a week or two ago that discussed various alternatives, I think ZFS with checksums, which unlike raid will not replicate corrupt data from drive nearing failure to the healthy drive.
They should have been able to swap out the RAID card and import the configs from the disks. This is pretty common practice with LSI/Avago-based cards. Curious as to why they didn't do that. Maybe the card went really bad and started writing garbage all over the place?
