Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Tonight, while working away on my PC, it started temporarily freezing. First, I'd hear a louder-than-usual click from the hard drive bay. Then, nothing would seem out of the ordinary until I actually clicked or right-clicked on something. Then, I'd get the little spinning ring (this is on Windows 7). The mouse would still move, but Explorer became unresponsive. Sometimes, I could use Alt-Tab to get to the app switcher, but it wouldn't actually switch apps. Then, after a pause of 15 seconds to a few minutes, everything would come back to life.

My strong suspicion was that this was due to a problem in one of the hard drives, but I couldn't tell which one (unless I wanted to take them all out, arrange them on the floor, and then surf the web while my ear was between them).

This is not the first time I've seen drives do this. But all of the tools I've ever used (Norton NDD, Spinrite, Ontrack, etc,) all seem to just try to read all of the sectors. But here's the weird thing: when I've seen drives fail like this, the first symptom is that they seem to, eventually, successfully perform the read (so the bad-sector scanner is happy). They just take 30 seconds to come up with a given sector.

Before I write a tool to read all of the sectors for errors AND anomalous delays, I figured I'd ask: does anybody know of a tool (for Windows, Linux, or on boot-cd) which can detect per-failure symptoms like this?

share|improve this question

Windows does some rudimentary logging of disk related failures. Take a look at your event log for timeouts:

enter image description here

Or other relevant information:

enter image description here

Typically the event source is disk or atapi in the System log.

Another alternative are drive benchmarking tools like HDTach. You can benchmark the drive and see if there's a significant drop in one part of the disk or another. It's not conclusive (it's hard to test these things fairly), but if we knew how to identify failing hard drives conclusively we wouldn't be here.

enter image description here

share|improve this answer

You can use the built in S.M.A.R.T. to get that kind of data, but the tools from the manufactures themselves are the best tools. They are also the ones you'll need to run (usually) before you can get an RMA.



I would suspect the OS drive first or whatever drive holds the swap data since you were getting the blue donut of patience.

share|improve this answer
SeaTools (last time I used SeaTools DOS) has an acoustic test that will spin down a drive then spin it back up, so you can use it instead of tak(ing) them all out, arrang(ing) them on the floor, and then surf(ing) the web while my ear was between them. – Jan 15 '12 at 12:27

You can just run Process Monitor for a while and just look for any abnormally long File operations.

Just filter it down to Query File or Create File operations.

share|improve this answer

You must log in to answer this question.

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