An open community 
of Macintosh users,
for Macintosh users.

FineTunedMac Dashboard widget now available! Download Here

Previous Thread
Next Thread
Print Thread
Retired Block Count, DriveDx reporting a problem
#49124 06/20/18 11:02 AM
Joined: Aug 2009
OP Offline

Joined: Aug 2009
I have an external OWC Mercury Electra 480GB SSD (that I use for Carbon Copy Cloner backups).

DriveDx v1.8.0 is now warning that the drive’s Retired Block Count is 32 and that the drive’s life-span may be detrimentally affected.

My problem is that I have no understanding and appreciation for what this portends. Is this block count issue normal for a two year old SSD, is the drive likely to have a catastrophic failure or will it gracefully lose capability, and about how much longer can I expect to use it?

I plan to call OWC tech support, but before doing so, I would like to have at least a rudimentary understanding of the issue and what I’m talking about.

All thoughts welcome.


Harv
27" i7 iMac (10.13.6), iPhone Xs Max (12.1)

Those who can make you believe absurdities can make you commit atrocities. ~Voltaire
Re: Retired Black Count, DriveDx reporting a problem
Pendragon #49125 06/20/18 01:27 PM
Joined: Aug 2009
Likes: 16
Moderator
Offline
Moderator

Joined: Aug 2009
Likes: 16
Every drive comes from the factory with spare data blocks, and when a data block on the drive goes bad that location is mapped to one of the spare data blocks. This is a normal part of drive operation and in fact it is common for a brand new drive to have retired data blocks. The retired block count is the number of data blocks that have been detected as bad. If the number of bad data blocks is growing that is often the result of the magnetic media flaking off of the drive surface. A study by Google Labs several years ago found that NEW bad data blocks are the single best indicator of impending drive failure.

Drives are rated with a MTBF (Mean Time Between Failure) and that is typically very long but there are always drives at either end of the statistical bell shaped curve. So while one drive may never fail in normal use statistically another of the same make and model may last only a few years or even months.

If I were you I would rerun the test periodically — say weekly or monthly — and watch the number of retired data blocks. If that number grows then replace the drive quickly.

Personally I always buy the enclosure and drive separately. That way I get to select a drive brand that has the best record of longevity — typically HGST or Toshiba. That also means I typically get a longer warranty on the drive even though I purchase both from the same vendor — typically OWC.


If we knew what it was we were doing, it wouldn't be called research, would it?

— Albert Einstein
Re: Retired Black Count, DriveDx reporting a problem
joemikeb #49127 06/20/18 01:52 PM
Joined: Aug 2009
Likes: 7
Offline

Joined: Aug 2009
Likes: 7
Considering that Harv has a SSD, does your statement about magnetic media flaking off have any relevance? I have the same SSD, by the way.


Jon

macOS 11.7.10, iMac Retina 5K 27-inch, late 2014, 3.5 GHz Intel Core i5, 1 TB fusion drive, 16 GB RAM, Epson SureColor P600, Photoshop CC, Lightroom CC, MS Office 365
Re: Retired Black Count, DriveDx reporting a problem
jchuzi #49135 06/20/18 02:56 PM
Joined: Aug 2009
Likes: 16
Moderator
Offline
Moderator

Joined: Aug 2009
Likes: 16
Originally Posted By: jchuzi
Considering that Harv has a SSD, does your statement about magnetic media flaking off have any relevance? I have the same SSD, by the way.

I missed the SSD part blush but the "retired block count" is the rough equivalent of bad blocks on a rotating rust media. There will always be some retired blocks which will occur as a normal part of operation as SSD data blocks are good for only a finite number of write operations. But the appearance of 32 would indicate the SSD is beginning to approach its end of life. There will naturally be an increase in the number of retired data blocks over the life of the drive, but I would think the rate of increase would be the best indicator of when to replace the drive.

Personally, given the nature of SSDs and how Time Machine works, I would have to think long and hard about the advisability of using an SSD for Time Machine backup. Speed of access is not really a major or even minor factor in Time Machine and SSDs do not have a long memory without refreshing and lots of write activity as is not recommended. Perhaps more important, high capacity HDDs are really cheap.


If we knew what it was we were doing, it wouldn't be called research, would it?

— Albert Einstein
Re: Retired Black Count, DriveDx reporting a problem
joemikeb #49144 06/20/18 10:37 PM
Joined: Aug 2009
OP Offline

Joined: Aug 2009
The short version...

Spoke with two OWC techs. After fumbling around a bit, the first tech called in the A team. The new tech opined that DriveDx often gives false reports with OWC SSDs. And that while 32 Retired Blocks is not a good sign, it’s difficult to tell just how bad, if at all, the condition may portend, and if the number of retired blocks begins to climb, then I should definitely call back as that would surely indicate a problem with the drive.

The senior tech had me use DiskUtility to check to the SSD as that often gives better/more accurate data than does DriveDx (well, according to him). But how true that is, I dunno.

At any rate, tech #2 seemed sincere, gave me a case number, and again encouraged me to call back should my situation worsen.

I now have an additional/different external drive installed, so for the interim, I am agreeable with waiting & watching. Still, I would be happier were OWC to replace the drive.


Harv
27" i7 iMac (10.13.6), iPhone Xs Max (12.1)

Those who can make you believe absurdities can make you commit atrocities. ~Voltaire

Moderated by  alternaut, cyn 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.4
(Release build 20200307)
Responsive Width:

PHP: 7.4.33 Page Time: 0.014s Queries: 24 (0.010s) Memory: 0.5917 MB (Peak: 0.6599 MB) Data Comp: Zlib Server Time: 2024-03-29 08:54:23 UTC
Valid HTML 5 and Valid CSS