Reply 48400 of 52819, by BitWrangler
- Rank
- l33t++
Yah I just let small amounts of randoms accumulate over long periods without worrying, if they cluster though, that's a surface damage indication, and if I desperately need to keep using that disk I will partition them out. (Put partition above and below and leave that "cylinder" or two unallocated and hopefully unaccessed) Probably I'll call something "done" when it's got randoms in the hundreds even if there are no spreading patches. Unless it's a unicorn drive, (XTs, early ATs, laptops with limited types etc) then it's just a case of running it and crossing your fingers, while not having any unique data kept on it.
I know this sounds ridiculous, but to use a "really spotty" drive, unless the CPU is hoplessly weak in it's class, I will make a compressed volume on the drive. Then scan disk it a few times... if errors keep happening, there's not much hope, but they typically limit to a couple of passes as the CVF settles in. But, what you've got here now is an extra layer of error detection/compensation/correction. This is all from experience by the way with being broke through the 90s when storage prices were insane capacities miniscule and I bought junk box hard drives for pocket change and tried to get use out of them.
Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.