PCA Forums

ive been getting delayed write errors on my boot drive. this is in a sys thats been running fine for 2 yrs with no changes. it also bsods w/o a code or just locks after the delayed write error. im thinking bad hdd. anyone have a second opinion?

Delayed can also be caused by a bad drive cable, dirty contacts on the drive cable, or a bad controller. Or yeah, the drive could be failing.

Re-seat both ends of the cable, or replace it.

Make sure you don't have an intermittent power connector on the hard drive. Might be a loose molex pin.

Then run the factory diagnostic. Run the advanced test.

checked the cables, they seemed ok. another hdd does share the controller it also seems ok. ran smart test: fail ran short dst: pass ran long dst: pass running spinrite l4. also this is a sata 3.0 gb setup if it makes a difference. and does spinrite wipe the drive at l4? i never used it before now.

If you're getting a SMART error then get your data off it while you can. When you get a SMART error it usually means the drive is on death's door.

No Spinrite does not wipe the drive. Level 4 "strengthens" the signal of each sector recording. It reads each sector, writing it back and inverting the bit order twice.

Yeah, save the info while you can. That thing is going down. May not be this week or month, but it's right when you forget about it.

spinrite says theres nothin wrong with it. but i backed it up with acronis anyway. if it croaks now i can just rma it i have a western digital that failed the smart test out of the box 2 yrs ago its still running ok. a fluke?

Be aware as I said in my Spinrite rant that just because Spinrite doesn't find an error, doesn't mean there wasn't a problem. Spinrite just forces the drive's own internal CRC routines to check every sector. The only time Spinrite becomes aware of a problem is if the drive tries to mark a sector bad and throw out the data in which case Spinrite tries to recover the data. Only at that point you will you see a "B" or "U" for a bad or unrecoverable block in the Spinrite disk map.

As for SMART, yes you can have a drive that fails SMART but passes all other tests and runs fine for years. SMART is just a set of statistics on the drive's performance (temperature, error rates, error correction rates, etc). SMART is tripped when any of these stats exceed the manufacturer's recommended limits.

So yeah the drive might be working, but it may be using error correction at a rate WAY beyond factory spec which is triggering SMART. The error correction may always work so the drive is basically fixing itself, but the drive is also exhibiting behavior that the manufacturer would consider outside of specification. This could also be causing performance issues as a side effect.

There's a freeware app called Active Hard Disk Monitor (the Pro version is not free) and it will show you all the SMART data along with a health rating for your drive. The problem with SMART is it usually tells you there's a problem after it's too late.

ok, sounds like i should trade them in while i still can then. the health on the seagate is 56% and the wd 88% the raw read error rate column on the seagate....value:113 worst: 99 threshhold:6

It's hard to read those values - there's no universally accepted set of numbers for them and HDD manufacturer's are VERY close lipped about what numbers they use.

But it's never good to see things over threshold.

If it's under warranty I would always RMA any questionable drive. No brainer.

that active hdd thing is nice, even tells total hours usage, anyone know if they still send refurbished replacement drives?

Yes they always send refurbs, and the refurb will have a warranty expiration the same as your original drive.

Nothing wrong with a refurb. It's been taken apart and put back together and shipped out with factory new tolerances.

got the replacement today, it also fails the smart test but the smart info displayed in active hard disk all reads 100% it also runs cooler and no bsods.