I have been happily running a RAID 10 of OCZ Intrepid 3700 pulled from an enterprise appliance. As soon as I updated to 5.7.2 all of them started reporting uncorrectable SMART errors (198) and that all the drives should be replaced. Other utilities show the SMART status as ok, so I am really skeptical that they are actually faulty.
Send a SoftRAID tech support file to support at softraid.
uncorrectable errors are unusual on SSD drives, but this is a state reported by the drive. It means it could not do the checksum, or the data was unreadable.
I agree they are probably not "faulty", but this is a fault condition.
I am unsure how to clear these flags ,which could have been caused by how the machine shut down, or how the enclosure powered down.
It is possible a 3 pass "Certify Disk" would clear this flag, but that means a backup/restore cycle.

