We have it resolved, but it was an issue with the RAID array, one we were away of and trying to fix but rebuilds were aborting because of the amount of bad sectors. It was reading off the other drive where sectors weren't as bad, even though the state was not consistent. We have managed to move the running copy onto a proper running RAID array and make it live, so that won't be an issue again when we restored the data back to the 1st data.