5

I have a Synology DS213 with two 600GB drives in RAID 1.

Last night the device reported that my second drive had become degraded and that I should replace it. When I ran a extensive S.M.A.R.T. test the results said that the drive is okay.

How can I confirm that the drive is actually bad? Is there any case that the degraded drive is the good one and that it is actually the other drive that is bad?

Nathan Villaescusa
  • 151
  • 1
  • 4
  • 5
    SMART tests are really not that useful. I'd remove the drive from the array and test it aggressively with the best disk drive testing program you can find, one offered by the manufacturer. If the drive passes the test, you can put it back into service in the RAID array. – David Schwartz Oct 23 '12 at 05:47
  • Also, you need to have "offline data collection" enabled in the drive, otherwise a lot of data will be missing. – Simon Richter Oct 23 '12 at 12:09
  • Microsoft research has found (via analysis from windows error reporting) that the odds of a second OS crash due to an HD error after a first is 1 in 3.5 for computers that have reported at least 30 days of system uptime. If you have a second crash the odds of a third are 1 in 1.7. Hardware failure is rarely a random one off event; if you care about the data on your system you should immediately replace the drive after the first error. http://research.microsoft.com/apps/pubs/default.aspx?id=144888 – Dan Is Fiddling By Firelight Oct 23 '12 at 13:00

2 Answers2

11

S.M.A.R.T. can be used as an indicator that there are drive problems but can never be relied upon to indicate that a drive is good. When there is disagreement between multiple diagnostic systems always favour the one that shows the worst results.

John Gardeniers
  • 27,262
  • 12
  • 53
  • 108
  • 3
    +1 for "When there is disagreement between multiple diagnostic systems always favour the one that shows the worst results.". – Soviero Oct 23 '12 at 03:09
1

As mentioned before, S.M.A.R.T. is great for alerting you when there are problems, but it definitely is not the catch-all solution for drive testing. I would remove the degraded drive and run tests against it to see what the issue is.

I recommend using SeaTools http://www.seagate.com/support/downloads/, there is a version you can run in Windows, or a bootable ISO if you want it to be OS independent.

The short test of SeaTools will do some basic checking, such as S.M.A.R.T. tests among others, but this will not guarantee to find errors on a drive, if they exsist. Running the long test will take a bit longer, but it will test each individual sector of the drive, and is MUCH better for finding out if a drive needs to be replaced or not. I've had many drives that have not had S.M.A.R.T errors and pass the short test, but failed on the long tests.

LandonWO
  • 216
  • 2
  • 6