>>3861724>>3861860This is normal if there was a hardware failure. Data recovery is rather specialized.
>>3861721The way you mitigate it is ahead of time. More copies, RAID6, stuff like that.
>>3862071Apart from spotting immediately abysmally bad drive series (which are rare, but I guess they existed a few times), there's not much you can gauge. Again, almost all drives will fall between more than 0% and 2% annual failure rates until they get "old" (whatever that exactly is, you'll never know in time - 6-10 years later there won't be many reviews until they stop making new drive models) and their failure rate goes up.
Again, the way to deal with this "safely" is basically SOLELY to have more drives. In RAID6/5/1 or as copies. Having more copies does very predictably and massively increase the chance of your data not dying to drive failure.
E.g. the chance of 2 or 3 simultaneous (like, within a week until you replaced it) drive failures is very, very low for a smaller number of drives that themselves only have a chance of failing 2% or less annual chance.
