0

I have a question about SSDs and bad blocks. I use a Mac and an acquaintance has an SSD installed on his Mac too. He used Scannerz (http://scsc-online.com/Scannerz.html) to confirm that the bad block was there. However, and this is the "big deal" if you will, it became a bad block after it had data in it. He had a video file that showed a demo for his company and it used to work fine. Then, all of a sudden, the thing would get to the exact same place in the movie, the player would go into spinning beach ball mode, and then eventually give up. A scan on the drive revealed a bad block.

SSD technology seems to be changing rapidly, but I think that if an SSD detects a bad block during a write, it re-maps it. What do they do if they detect a bad block during a read, meaning one that, at one point in time, just decided to go bad even though it was containing data? It looks to me like they just leave it in place. If they didn't it would destroy what was left of the file and I think it would make a file completely unusable. As is, this is at least partially recoverable.

Does anyone know?

Thanks.

1
  • I can't speak to SSDs, but I know for certain that the vast majority of HDDs work this way. They only remap bad blocks on write, not read.
    – Spiff
    Mar 31, 2014 at 23:50

1 Answer 1

0

Your friend is experiencing an I/O error because the SSD can't read one (or more) bad blocks. The bad block is corrupt. HDs and SSDs are not the same, but they handle errors similarly.

When a hard drive gets a bad sector, it's usually due to a head crash of some sort that physically damages the media rendering the sector or the seek traces on the media unreadable. A hard drive can re-map these during a write operation to spare sectors if they're available. During a write operation is the only time the controller attempts to correct them.

With an SSD a marginal block may "just go bad" for no apparent reason. It's usually a manufacturing defect. There are no moving parts in an SSD so physical damage (except in extreme cases, like dropping it off the Empire State building) isn't likely. Like an HD, most, if not all correct this during a write operation. The logic is to leave the file in place so that as much that is recoverable can be recovered, which make sense.

Unfortunately with SSDs, being an essentially emerging technology, how one manufacturer and another manufacturer handle errors are totally up for grabs. I've seen some controllers leave bad block in place and writable until they're allowed to go through their clean up routines (a BAD controller) while others will immediately exclude that block from use and properly re-map it. The latter is the model that seems to be gaining acceptance, as it should, because you shouldn't be writing data to a bad block.

As I said, SSDs are still an emerging technology, but they aren't and never will be perfect, regardless of what marketeers tell you.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .