Have I inadvertently fixed my Samsung HDD?

A drive (Samsung 2TB HD204U) in an external enclosure was acting up at the end of last week. So I checked on HD Tune Pro and saw an issue with Current Pending Sector which was steadily increasing in value as more data was written (FYI - drive was v. full). So I did the sensible thing, checked it out with Seatools and it failed on a Short Generic Pass. Next step, as it was in warranty still with Seagate, was to set up an RMA. Then I swapped it out for my backup and then copied the backup onto a fresh drive. All good so far.

Now whether I'm RMA-ing a drive or selling it on I always format and then do an erase of some sort. Being a 2TB hooked up via USB2 I went for a unused space 2 pass Russian GOST P50739-95 as I didn't really want to wait the best part of 4 days to complete!

Over the latter course of the erase (using Eraser 6) something interesting began to happen, the value of Current Pending Sector started to drop from 48 to the teens and then finally to zero. So I started to wonder if the erase had fixed the drive. On rechecking Seatools, using a couple of the tests, it received a pass. This leaves me wondering whether I have effectively fixed the problem as the problems sectors have been successfully overwritten. I'm no hardware expert but I have some doubts, but I'm in a tricky situation with the RMA now as the drive appears to have a clean bill of health!

Can someone advise what the best course of action may be?

Thanks.

The only thing I have to offer is that it's never a good idea to get a drive "very full." It looks for space for its read/write functions and when it gets full that creates performance issues. So it COULD be that it was just too full, but I don't know anything about "Current Pending Sector" stuff, which may point in another direction (or maybe being too full could have caused that too, I don't know).

The other thing that can cause a drive to "act up" is overheating. I have a 1.5TB in an aluminum external case (no fan) and I thought at one point it was going south because of noise it was making... but then realized it was very hot in the room and the drive might just be hot. I brought a floor fan over and directed it to the case, and the drive started behaving again and hasn't given me a problem since, some 2 yrs later. But when it's hot (outside, I mean) I always put a fan on it when running large jobs.

Maybe someone else can help you more...

Yes, it certainly is way too full. I'm going to be doing a fresh install with W7 HP 64bit shortly which will allow me to start using 3TB drives, so there will less need for data juggling between 2TB externals.

As for heat, the drive was pretty cool running in an Antec MX-1 (sadly they don't make them anymore for some reason). I think you'd be hard pushed even on the warmest of days to see it reach 30C. So it's unlikely to be that.

Wikipedia defines Current Pending Sector as 'Count of "unstable" sectors'. I'm not entirely sure I fully get the explanation that follows, so I'm not a great deal wiser whether the drive is safe to re-use. I feel I'm a bit stuck inbetween not safe to use and no longer able to RMA. However I have just discovered this thread which details the same issue. . .

Current pending sector count reset - AnandTech Forums

That heartens me a bit since my drive hasn't had any reallocation events.

I think I'll re-run Seatools but this time do the Long Generic test and see if it snags something. Whatever happens I'll contact Seagate to see where I stand.

DangerFourpence said:

I think I'll re-run Seatools but this time do the Long Generic test and see if it snags something. Whatever happens I'll contact Seagate to see where I stand.

The long test is a must as it might find issues the short test doesn't, and it's likely what Seagate would have you do before RMA'ing. IAC if it finds errors you have a leg to stand on to return it.

As for going back to square one (assuming the test comes back fine), my personal philosophy in troubleshooting difficult things is to eliminate every factor it COULD be to see if the problem persists. In your case the full drive could cause issues, so if you can't return it or decide not to, you might find it runs perfectly fine as long as you don't stuff it to the gills. With my 1.5TB when it got to only having about 235GB of space left, I promptly started transferring data off to another drive, not wanting to get anywhere near the threshold of problematic behavior. As it is that's about 85% full, which is full enough, IYAM.