4

So, my hard drive was acting all slow, and after checking through my syslog, I found all these errors. Some googling led me to try to run e2fsck -c -c on the drive. So, it's been going for about 2 days now, and seems to be nearly complete. But, it says this:

e2fsck 1.42 (29-Nov-2011)
Checking for bad blocks (non-destructive read-write test)
Testing with random pattern:  92.64% done, 43:38:19 elapsed. (97/0/0 errors)

That's cool. But, what does the 97/0/0 signify? How much data loss can I expect? (It's a 3TB hard drive) (I guess I could just wait for ... at this rate, I guess 3 and a half hours, but I'm not sure if it'll tell me what the 97 means, so I figured I'd ask.)

Thanks!

bnsh
  • 143
  • 4

1 Answers1

5

That means there was 97 read errors, 0 write errors and 0 corruption errors while testing your disk with random patterns.

I wouldn't trust that much a disk with read errors.

jlliagre
  • 8,691
  • 16
  • 36
  • Interesting. So, you'd say I should copy as much as possible from there and replace the drive, even if e2fsck marks those blocks as "bad"? – bnsh Jan 12 '13 at 17:17
  • Yes. Even if those blocks are marked as bad and not used anymore, because bad sectors tend to follow the bathtub curve. (Read: If you get a few now, then there is a decent chance that tomorrow you will have more then a few). – Hennes Jan 12 '13 at 19:24
  • OK, cool... (Well... not so cool for me, since I have to get a new hard drive, but, cool that I have a definitive answer at least!) Thanks! (I'd upvote, but I'm too much of a newbie! D-oh!) – bnsh Jan 12 '13 at 19:39