Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

savoirfaire

macrumors 6502
Original poster
Nov 23, 2003
353
38
New England
Hi, all -

Last night, I ran DriveDX to determine the hours-used on a SDD that I was selling, and I incidentally discovered that the SMART status of my 3TB media drive is being reported as "Failing" with an overall health rating of "BAD 1.4%".

The reason for this is apparently a seek error rate of 31. How bad is this, really? Searching on Google suggests that I best get my data off of there, and the sooner the better, but some people say not to read too much into it. In any case, the drive is out of warranty already, so I've ordered a new 4TB drive and plan to transfer all my data over on Tuesday...
 
That's pretty bad.

Theoretically, the seek error rate has nothing to do with data integrity. The actual storage medium (the disk drive platters) are fine, but for whatever reason the mechanical system that drives the read heads is degrading in some manner.

As far as the actual SER reading goes, the lower it reads the worse the issue becomes. Most drives (depends on manufacture) start out at 100, which lowers as the number of errors increases. 31 is roughly means that every 1/1000 seek attempts has failed, and the drive had to correct the head positioning accordingly.

Your drive will do hundreds of thousands (millions?) of seeks per day. If 1/1000 seeks is failing, that's a pretty high error rate and I would say that your drive is definitely failing. When (not if) the head mechanism permanently fails, your data will be inaccessible unless you send it into someone who can perform platter level recovery (but this will cost you $$$$$).

TLDR; get your data off that drive ASAP.

-SC
 
That's pretty bad.

Theoretically, the seek error rate has nothing to do with data integrity. The actual storage medium (the disk drive platters) are fine, but for whatever reason the mechanical system that drives the read heads is degrading in some manner.

As far as the actual SER reading goes, the lower it reads the worse the issue becomes. Most drives (depends on manufacture) start out at 100, which lowers as the number of errors increases. 31 is roughly means that every 1/1000 seek attempts has failed, and the drive had to correct the head positioning accordingly.

Your drive will do hundreds of thousands (millions?) of seeks per day. If 1/1000 seeks is failing, that's a pretty high error rate and I would say that your drive is definitely failing. When (not if) the head mechanism permanently fails, your data will be inaccessible unless you send it into someone who can perform platter level recovery (but this will cost you $$$$$).

TLDR; get your data off that drive ASAP.

-SC

Aye aye, Captain! Thanks for that confirmation, most of what I read did sound pretty bad. Actually, what was amusing is one guy asked about it on HardForum (http://hardforum.com/showthread.php?t=1681648) and someone said not to read too much into seek error rates and such. Two weeks later, he posted back that the drive died!

Good thing I stumbled across that when I did. I just hope it survives long enough for me to transfer off my data! *gulp*
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.