MacInTouch Reader Reports

Hard Drives: SMART

Feb. 22, 2010
May. 27, 2010
Aug. 30, 2010
Aug. 31, 2010
Sep. 4, 2010
Sep. 7, 2010
Sep. 8, 2010
Sep. 9, 2010
Sep. 10, 2010
Sep. 11, 2010
Sep. 27, 2010
Sep. 28, 2010
Sep. 29, 2010
Nov. 6, 2013
Nov. 8, 2013
Nov. 9, 2013
Nov. 11, 2013
Dec. 9, 2013
Dec. 10, 2013
Mar. 19, 2014
Mar. 20, 2014
Mar. 21, 2014
Mar. 22, 2014
Oct. 3, 2014
Oct. 12, 2014
Oct. 13, 2014
Oct. 14, 2014
Feb. 2, 2015
Feb. 3, 2015
Jun. 11, 2015
Jun. 20, 2015
Jun. 22, 2015
Jun. 23, 2015
Jun. 24, 2015
Jun. 25, 2015
Jun. 26, 2015

Newer entries...
Feb. 22, 2010

item.109948

WK Lee

I've been looking at portable storage devices and while most of them don't implement SMART, I noticed that Hyperdrive implements it in their enclosures.

http://www.hypershop.com/HyperDrive-COLORSPACE-UDMA-Casing-Only-p/hdcsu-000.htm

This is a brand new product:
http://www.hypershop.com/ProductDetails.asp?ProductCode=HDA-000

I don't know if the information is passed through USB and would be recognized by a program like SMART Utility or not.

May. 27, 2010

item.115138

MacInTouch Reader

I've been working with the command line utility "smartctl" and have recently noticed that a small number of the new computers shipping from Apple have the SMART support disabled. I get the following message:

SMART support is: Disabled

I can easily enable it and it then functions normally.
So far, I have only noticed this on Hitachi drives.

Aug. 30, 2010

item.120119

David Minehart

I have a new Mac Mini 2010, and was trying out the new SoftRaid 4 on it. SoftRaid reported that the internal Hitachi 320GB drive showed excessive bad sectors, and was 20-60 times as likely as a good drive to fail within the next 2-6 months.

On the SoftRaid site, I saw mention of their using the results of research done by Google on predictors of drive failure, but no details.

Anyone have a better idea of what's going on here? And whether there's a utility I can use to get data to present to Apple, if warranted, to see about getting a replacement drive?

Aug. 31, 2010

item.120145

Colleen Thompson

David Minehart asked

"...SoftRaid reported that the internal Hitachi 320GB drive showed excessive bad sectors, and was 20-60 times as likely as a good drive to fail within the next 2-6 months...Anyone have a better idea of what's going on here? And whether there's a utility I can use to get data to present to Apple, if warranted, to see about getting a replacement drive?"

I know I sound like a stuck record here, but SMART Utility from http://www.volitans-software.com will read an internal disk's SMART attributes directly and tell you if there are any bad blocks (sectors). It doesn't work with external drives (USB/FW) in most enclosures. There is a free trial period but it's well worth paying for. I know there are command-line utils that are free but this is extremely easy to use for those of us not so familiar with Terminal.

My only connection with Volitans is as an extremely satisfied customer.

Sep. 4, 2010

item.120477

David Minehart

Well, I located the free smartctl command line utility, installed it with some effort, and am happy with the output. The 30 minutes of setup are worth the $25 for the commercial product that puts a GUI face on the same built-in command. What I didn't locate was a way to be confident that the report was accurate. Since it didn't recognize the specific drive I was using, I had to let it use "permissive" techniques to read the data. So, maybe it's not reading the drive's performance correctly.

I decided to just return the new Mac Mini to Amazon and get a replacement. Two hours after I took a minute on their web site to request a return I got an email saying the replacement had been shipped! I couldn't have asked for better service.

Sep. 7, 2010

item.120541

François Schiettecatte

You can also use SMARTReporter to monitor the SMART status of drives. It uses smartctl and will give you output from that in a text document. And it's free.

item.120613

Colleen Thompson

Francois Schiettecatte said

"You can also use SMARTReporter to monitor the SMART status of drives. It uses smartctl and will give you output from that in a text document. And it's free."

I do have, and use, SMARTReporter (unlike SMART Utility, it will email notice of failure.)

The SMARTReporter FAQ contains this:

Q: Why does SMARTReporter report another S.M.A.R.T. status than SMART Utility, TechTool Pro, ...?
A: SMARTReporter (like Apples Disk Utility.app) only looks at the S.M.A.R.T. summary status, which is an overall value calculated (by the drive) from the individual S.M.A.R.T. properties...
As far as I know, SMART Utility and TechTool Pro have mechanisms to calculate their very own summary status from the individual properties itself - ignoring the summary status S.M.A.R.T. provides - whether this really provides superior error-prediction is not known.

Well, it's known to me. The whole problem with the summary status in Disk Utility and other apps is that the thresholds for failure are way too high, especially when reporting bad blocks (sectors), which is the type of failure I see most often. I've had disks that were entirely unusable that DU confidently proclaimed as "SMART Status Verified."

Yesterday I had a G5 tower whose drive had 3500 bad blocks (and it wouldn't boot), and DU actually reported the drive as having failed. But I've had other ones with dozens of bad blocks that were reported as fine by DU. Knowledgeable people usually replace a drive which has *any* block failures; while you can do a Write All Zeroes reformat to exclude the bad spots, that kind of failure is almost always a continuing process and it's better to replace the drive than to take chances with your data.

I see that now SMARTReporter will give you direct access to the values of individual attributes (but only on Intel Macs). I'm on a PPC Mac right now so I can't see what it reports, but that is a useful feature in terms of checking for bad blocks. I have found SMART Utility's own summary very reliable and useful in predicting or confirming drive failures. (You can also use free command-line utilities.)

Sep. 8, 2010

item.120623

Olve Peersen

I'll just reiterate that in my experience, SMART Utility has a much better track record of diagnosing drive problems that SMARTreporter or Disk Utility. My impression, which I admit is now a few years old, is that SMARTreporter more or less regurgitates what the drive is indicating, much like Disk Utility, but it does have the automatic email notification feature. SMART Utility, on the other hand, does a much more aggressive analysis of the smartctrl logs and then give you a simple passes/failing/failed "grade" at the end.

I have now had many systems that were active a bit finicky (slow, crashing...) where the ultimate fix was a new hard drive - often with a cloning of the old one. Disk Utility and SMARTreporter did not indicate drive problems, but SMART Utility did show a failing status. It's hit rate is essentially 100% in my experience and on old systems it is first in line of my diagnostic tools.

item.120629

Brian Marsh

SMARTReporter as of version 2.5.0 now checks for I/O Errors in the system.log once every 10 minutes
(the recent Sept 5 release had to disable this feature for PPC macs after finding some kind of bug related to this check)

This is a nice addition, and while limited to internal drives only (it seems, I should be able to confirm this within a few days here at work) it may pick up drive failures that SMART alone will not.

item.120654

Thib Hal

In my experience the SMART status has been of absolute no use to me.

My original MacBook Pro had a hard drive failure and it had to be replaced with a new drive. That's when I decided to install a SMART status tool to make sure I was informed ahead of time.

When I bought my MacBook Air, I installed one of those SMART status tool things too. Well, this MBA seemed to be having some hard drive issues after about 2 years and I ran Disk Utility and everything seemed fine. I checked the SMART status log and everything seemed fine. One day, I happened on a different utility (forget which one that was) and it informed me that there were bad blocks in my hard drive. Well, I reformatted the drive and I guess there were enough good and free blocks left so the utility no longer reported bad blocks. A while later, this same utility reported bad blocks started appearing again. SMART status did not report a darn thing. I tried to reformat the hard drive and that didn't work.

I brought the computer to an Apple store and they replaced the drive for free. SMART status in all these cases were of absolute no use to me.

Since having a new hard drive installed on my MacBook Air, I have removed the SMART status reporting utility. It seems to do me no good and only provided a false sense of security.

SMART a'int that smart. It's more like erm... "SMART-ASS" ;-)

Sep. 9, 2010

item.120680

David Weatherston

Automatic SMART reporting didn't catch a recent problem with the drive in my iMac - it was the slight and sudden squirrelliness of the drive that alerted me to a developing problem. What SMART did in this case was provide immediate certainty that there was a relatively minor hardware problem. Accordingly I cloned the drive promptly (thank you, Mr Bombich) and had the machine serviced, thus avoiding all the gnashing of teeth and rending of garments that attended a drive failure some years ago.

I only wish that SMART could be made available for external drives.

Sep. 10, 2010

item.120807

Colleen Thompson

Thib Hall complained that SMART ain't smart.

However, what Thib's really complaining about is some utilities'
*interpretation* of his drive's SMART attributes, though Thib does not say which utility was in use, and then goes on to say that yet another utility did show bad blocks.

To repeat my earlier comments, Disk Utility, SMART Reporter, and others that use the standard thresholds to assign Pass/Fail to a drive will not alert you, in many cases, until the drive is falling off the cliff of disaster. You can either learn to interpret the attributes yourself, or use something like SMART Utility which has its own interpretation algorithm, and which I, personally, have found to be almost 100% accurate in predicting or confirming a failure.

(Note the "almost". I've had 2 failed drives from unknown causes where there were no SMART attribute errors.)

Sep. 11, 2010

item.120906

Matt P

In the last year, I've had two WD drives go bad. One was a 250gb and had mechanical failure and the other most recent was a 1.5TB "green" drive that reported SMART errors almost from the beginning in Disk Utility. I ran the SMART Utility on the latter, it found bad smart status with 580+ bad blocks. Diskwarrior also reported it as a failed drive. They were both EARS series drives if that matters. One lasted less than two months, the other four months. Of course, Newegg will do nothing to take a drive back after 30 days so if I want my ~100 bucks back I have to deal directly with WD and get a replacement - and pay shipping for - of the same junky drive. I will never buy a WD again, I had had good luck with them til the last year. I also have a friend who had run WD's on servers up til about 2 years ago when WD quality seemed to take a huge nosedive and he lost several drives within about a year, all WD drives. Regardless of how good they might be about replacing them and honoring their warranty, who wants to deal with the hassle and data loss. Junk is junk no matter how you look at it. I'm thinking of going to Hitachi or Samsung (No thanks to Seagate, they are cheap in price and cheap in quality and have a horrible track record on Macs in the last 4 or 5 years) drives from now on.

item.120922

John D. Broughton

The thing about all platter type Hard Drives is that they're spinning, motor driven, mechanical devices that, in most cases, run under demanding loads in less than perfect places for long periods of time. Being mechanical, they are prone to fail at any time and will most definitely fail at some point in time. Murphy's Law states that if a hard drive is going to fail, it will do so at the most inopportune time and place. The only solution to this problem is to backup your data and backup your data frequently.

Sep. 27, 2010

item.121738

Sugi Fox

Yesterday I received a new 240GB OWC Mercury Extreme Pro for use as the boot drive. It works fine so far, but when I checked its SMART status today I got contradictory results:

SMART Utility reported 6272 reallocated bad sectors,
SMART Reporter signaled "S.M.A.R.T. condition not exceeded, drive OK", as did Disk Utility.

Does anybody know if SMART status and criteria are comparable for HDDs and for SSDs? Is SMART Utility's alert reason for concern?

Thanks for any informed opinion.

Sep. 28, 2010

item.121861

Steven james May

Sugi Fox wonders about SMART and the new 240GB SSD from OWC...

According to my rep at OWC, all of their SSDs should behave the same. I have installed 2 60GB OWC SSDs as System Drives and neither SMART Utility nor SoftRAID 4.01 reports SMART staus. SoftRAID shows the drive but reports the SMART status as "unavailable". SMART Utility does not see the OWC SSD at all-- it sees just the five internal 1.5TB Barracudas and reports on them.

Sep. 29, 2010

item.121907

Matthew Butch

re: SMART. I have noticed that SSDs can be very weird with SMART. Some of the information can be accurate, some isn't. We're still investigating exactly what the manufacturers are doing with SMART on SSDs.

Note: I'm the developer of SMART Utility.

Nov. 6, 2013

item.181592

Samuel Herschbein

MacInTouch has had several threads about Disk Utility and S.M.A.R.T. status. Here's another data point: for a failing original Seagate drive in a 2.5yo iMac that won't boot due to disk issues, Disk Utility continues to report the S.M.A.R.T. Status as Verified, while SMART Utility shows an increasing number of errors, including 23 Pending Bad Sector errors and over 10,000 other errors.

A great feature of Carbon Copy Cloner is that it makes a list of the files which encounter disk errors during copying. Very useful!

Nov. 8, 2013

item.181689

Peter Engels

Samuel Herschbein said:

MacInTouch has had several threads about Disk Utility and S.M.A.R.T. status. Here's another data point: for a failing original Seagate drive in a 2.5yo iMac that won't boot due to disk issues, Disk Utility continues to report the S.M.A.R.T. Status as Verified, while SMART Utility shows an increasing number of errors, including 23 Pending Bad Sector errors and over 10,000 other errors.

A great feature of Carbon Copy Cloner is that it makes a list of the files which encounter disk errors during copying. Very useful!

Two years ago, the HD failed in my iMac; neither Smart Reporter nor Disk Utility reported any problem until the day it actually failed. However Carbon Copy Cloner started reporting disk errors about ten days earlier. I emailed CCC about this, and got an immediate reply expelling what was happening, and telling me to get a new HD soon. So the morning that the actual failure happened, I had an appointment with Apple, had a good backup in hand, got a replacement HD and had no loss of data. I have been using CCC for backup for some years now, and without their support I could have lost it all. So use CCC for backup, and if it starts reporting errors, act immediately -- do not wait for Smart Reporter or Disk Utility to agree you have a problem.

Nov. 9, 2013

item.181770

MacInTouch Reader

I think this has been said around MacInTouch before, but SmartReporter is late to report issues.

When I was running it, my drive was not setting it off, even though I was having obvious issues with speed, which is often a good sign that things are failing. I then found SMART Utility, which immediately condemned my drive. It is not free, but it actually works.

I also use Carbon Copy Cloner and highly recommend it.

item.181772

MacInTouch Reader

Samuel Herschbein and Peter Engels reported that CCC's error reporting is useful as an early harbinger of drive failure, and Peter advised:

"So use CCC for backup, and if it starts reporting errors, act immediately -- do not wait for Smart Reporter or Disk Utility to agree you have a problem."

Just the other day I had a similar experience, but with SuperDuper.

While doing my regular nighty backup, SuperDuper reported a generic disk I/O error copying a file and stopped the backup job. Part of Shirt Pocket's recommended troubleshooting was to try to duplicate the file using Cmd-D in a Finder window. If Finder duplicated the file with no further error, the error was probably just-one-of-those-things, but if Finder reported an error duplicating the file, this could mean Bad Things. Oops, Finder reported "Error -36, the file could not be read or written," and gave up copying the file. The error persisted when I tried to copy the file to another external disk, eliminating the possibility that the backup disk was faulty, and pointing to a problem with the Mac's internal disk.

Long story short: The next day at the Genius Bar, their hardware test quickly reported that my Mac's hard drive had developed bad sectors and was failing, and I needed a new hard drive. Luckily, I had presence of mind enough to stop using the computer immediately upon getting the error, and booted from a SuperDuper clone backup, created a custom copy script excluding the offending file, and completed backing up my Mac's disk with no further errors. (SuperDuper is great.) It appears that the damage was limited to one file.

So, the morals of the story are: As Sam and Peter advise, errors from backup apps should not go unheeded, and Error -36 reported by Finder means the hard drive is starting to fail. Disk Utility and even Disk Warrior will not detect bad sectors (I ran both, and neither found any errors), because these apps only check the file system, not the physical disk. And, Disk Warrior's hourly SMART monitor test did not fire any alarm.

I am looking for an app that tests disks' surfaces to test my disks periodically; Drive Genius seems to be a leading candidate. (Any recommendations?)

It is unfortunate that Apple's revamped support pages are pitifully lacking in troubleshooting information, but you can print a copy of "Common System Error Messages: What they Mean and What Might Help Resolve the Problem" and bring it with you to the Genius Bar. (I found that only after a few hours of searching.)

(The machine is a one year old Mac Mini 5,2 with the stock 500GB, 5400rpm hard drive. Unfortunately, I never noted the drive's manufacturer.)

[SoftRAID offers a thorough testing/certification procedure among its features. -Ric Ford]

Nov. 11, 2013

item.181798

David McLaughlin

I strongly agree with Ric's recommendation of SoftRAID. I use a 4-pass run with the Certify feature on any new drive, including flash drives, before using them. You can do this even if you don't use SoftRAID to format the drive after certifying it.

One big advantage of a SoftRAID volume is that you can configure things to send an email in the event of trouble. I have received emails from my Mac at home twice indicating read errors and a failing drive. Both times confirmed by SmartUtility that the drives had bad blocks.

The only downside of a SoftRAID volume is that they cannot be "safe booted", but that's why you have more than one bootable volume right?

Yes, SoftRAID is expensive, but their tech support is top-notch (usually responding in minutes), and it has more than paid for itself by identifying problems before they became critical.

item.181835

MacInTouch Reader

On Nov. 9 I wrote of my experience with SuperDuper's generic I/O error and Finder's "Error -36" as harbingers of a failing drive, and I noted:

(The machine is a one year old Mac Mini 5,2 with the stock 500GB, 5400rpm hard drive. Unfortunately, I never noted the drive's manufacturer.)

When the Apple Store called to tell me the Mini was fixed with a new hard drive, I asked who the manufacturer of the failed drive is. The failed drive is made by Hitachi. Odd, because I've had nothing but good results with Hitachi drives. I guess it catches up with everyone eventually.

Dec. 9, 2013

item.183320

MacInTouch Reader

I have a weird situation. On my Mac Pro, running Mavericks, one of my other internal hard drives failed the SMART test, and [I'm] in the process of replacing it. My question is: Would that hard drive make my Mac Pro hang while accessing the boot drive and seeing too many spinning beach balls? Thank you.

item.183337

Stephen Hart

MacInTouch Reader wrote:

"Would [a failing] hard drive make my Mac Pro hang while accessing the boot drive and seeing too many spinning beach balls? Thank you."

At least in the past, Macs could hang while polling internal drives, even if the boot drive was good.
I don't recall this ever happening with externals.

Dec. 10, 2013

item.183359

Grandy Pollo

A MacInTouch Reader asked:

I have a weird situation. On my Mac Pro, running Mavericks, one of my other internal hard drives failed the SMART test, and [I'm] in the process of replacing it. My question is: Would that hard drive make my Mac Pro hang while accessing the boot drive and seeing too many spinning beach balls? Thank you

Assuming it is not the boot drive (and it sounds like it is not), remove it (Mac Pro = 2 seconds work) and find out. It could very well pose an issue if hardware is failing on being polled at boot time.

item.183360

Colleen Thompson

Re:

My question is:
Would that hard drive make my Mac Pro hang while accessing the boot drive and seeing too many spinning beach balls?

Yes. If it's connected, it can gum up the works.

At least in the past, Macs could hang while polling internal drives, even if the boot drive was good. I don't recall this ever happening with externals.

I've got a little external USB drive right here that's currently doing that. I don't really see why externals should be any different.

item.183369

Todd Bangerter

For MacInTouch Reader: yes, a failing hard drive could be directly responsible for an increased occurrence of spinning beach balls. If you open the Console application, and do a search in the system logs for "I/O error" and you see a bunch of entries like this:

10/31/13 6:12:06.000 PM kernel: disk0s2: I/O error.

then that means there was a period of time where the OS was trying to access data on, or write to, the indicated disk, and there was some kind of error (such as a failing hard drive) that caused the operation to time out. During this period when the OS is waiting for a disk response, it can often manifest as a spinning beach ball.

item.183386

Steven May

MacInTouch Reader wrote:

"Would [a failing] hard drive make my MacPro hang while accessing the boot drive and seeing too many spinning beach balls? Thank you."

Any malfunctioning drive can hang a Mac. And 2 failing drives at the same time will make you pull your hair out.

item.183398

Skot Nelson

Re:

Would that hard drive make my Mac Pro hang while accessing the boot drive and seeing too many spinning beach balls?

This is exactly what happens when my drives start to wear out. Replace it sooner, for peace of mind, rather than later.

item.183402

MacInTouch Reader

Yes. I have seen it many times. Need to disconnect the drive to make the other one work properly again.

Mar. 19, 2014

item.188023

Kevin C.

I recently had two hard drives begin to fail (a pair of drives in a RAID 1 mirror). Thankfully I had a backup and noticed it before the drives were completely dead.

If I had been running a third party program that monitors SMART I know I would have been alerted sooner and avoided some downtime. Time to make that investment.

Smart Utility by Volitans has been recommended for years and I've run the demo from time to time. But has anyone compared it to similar options like DriveDx by Binary Fruit or SMARTReporter by CoreCode?

I compared demos of the first two recently, and they seem to report identical data. DriveDx has a much more modern interface that includes more detailed descriptions in the tool-tips. It is also cheaper for individual and family licenses.

I'd love to hear thoughts from anyone else who has used several of these options.

Mar. 20, 2014

item.188072

Steven Klein

Kevin C. asked about SMARTReporter by CoreCode.

First let me say that I'm a Mac consultant; I've replaced hundreds of bad hard drives, and the vast majority that fail never generate a SMART error. In other words, it generates lots of false negatives.

(Does it also generate false positives? I don't know, but if I do see a SMART error, I always replace the drive.)

With regard to SMARTReporter: In addition to monitoring SMART status, it also monitors the logs for I/O errors. This feature has definitely saved me at least once. It alerted me to I/O errors, and I scanned the disk with Drive Genius and found bad blocks.

I see DriveDx also offers I/O monitoring, but I haven't actually used it.

I did download the demo version of Smart Utility, but it relies on a drive's SMART sensors, and experience has taught me not to put my trust in those sensors.

item.188052

James Cutler

Kevin C. asked about Smart Utility,

"... has anyone compared it to similar options like DriveDx by Binary Fruit or SMARTReporter by CoreCode? ..."

In my experience, the most common SMART data not shown by Disk Utility that indicates impending failure is reallocated sectors in the drive attributes. A small non-zero count here is still "Verified" by Disk Utility. It means "Replace this disk" to me. The ease of accessing drive attributes is of primary interest to me.

All three programs gather the same SMART data. All use the same SATSmart KEXTs to enable monitoring of external drives. All cannot monitor external drives if the drive enclosure does not properly support pass-through of SMART attribute data.

I find the user interface of DriveDx and SMARTReporter to be too "helpful". Here's why I think that:

DriveDX spreads out the information vertically, resulting in slower visual scanning, and offers bright colored status bars with confusing percentages (100% means 0, for example). The selection of parameter sets (Pre-Fail, Life-Span, etc.) is nice, but the status selection resets with each new device selected.

SMARTReporter's complex interface requires far too many mouse clicks to view drive attributes. If you have only one drive to check, this is OK. For more than one or two drives, it becomes painful.

Smart Utility has a lean and functional interface. All the most interesting data is displayed as each drive is selected in the list in the left pane. Each information block for the drive may be expanded for detailed examination without removing the master window from display. Problem drives are highlighted in red in the list making fast work of drive review -- the number of data points to scan is the number of drive identifiers in the list.

All three programs accurately present SMART data, but, for my money, Smart Utility is much more efficient in using my time to check multiple drives.

item.188088

Marc Heusser

About detecting hard drive failures with SMART...

When oh when will we get ZFS on a Mac, which would inform you of any read or write errors when they happen first, irrespective of where they happen? My number one wish for the next Mac OS!

item.188104

Colleen Thompson

I agree with everything James Cutler said, with one small correction.

In my experience, the most common SMART data not shown by Disk Utility that indicates impending failure is reallocated sectors in the drive attributes. A small non-zero count here is still "Verified" by Disk Utility. It means "Replace this disk" to me. The ease of accessing drive attributes is of primary interest to me.

I would correct "a small non-zero count" to read "a huge non-zero count".

I have seen drives with hundreds or thousands of bad sectors receive passing SMART status in Disk Utility.

Mar. 21, 2014

item.188119

Jeff Bagby

Kevin C is seeking feedback on SMART utilities for his RAID 1. I recommend using SoftRAID. In addition to the more in-depth SMART reporting and analysis offered by SMART Utility, SoftRAID also makes Kevin C's earlier posting about splitting a RAID 1, rebuilding a RAID 1, and even using an external drive as a third drive in a RAID 1 set a piece of cake.

Just a satisfied customer.

www.softraid.com

item.188150

James Cutler

Colleen Thompson wrote,

" ... I have seen drives with hundreds or thousands of bad sectors receive passing SMART status in Disk Utility."

In most cases, I don't see huge number of bad sectors because I monitor often and replace any drive that begins to show that symptom. Colleen certainly reinforces the case for using SMART Attributes as providing more diagnostic data then Disk Utilities "Verified".

Mar. 22, 2014

item.188176

Tracy Valleau

I have (and use) SoftRAID 4.5.3, and I'm hoping Mr. Bagby can show me where / how to get SoftRAID to show "more detail" than SMART Utility.

I've not been able to find it, nor is it in the user guide, as far as I can see. In fact, the SoftRAID documentation says the opposite: that it check only four of the SMART reporting points (since those are the ones found to point to drive failure.)

But in terms of "more detail" I seem to have missed how to get to it using SoftRAID.

Honestly, in terms of detailed SMART reporting, I've yet to find anything better than SMART Utility, particularly if you remember to use the "more info" buttons...

... but I'm always eager to learn!

Oct. 3, 2014

item.197554

Steven May

One of my servers is a G5/2.7MP. Attached to the 2 internal SATA ports are 2 WD raptors.

Additionally I have 3 Sonnet 8-port PCI-X SATA cards, two with 8 ports external and one card with 4 internal and 4 external ports.

I have 24 drives attached to the 3 Sonnet cards, 20 external drives mounted in 4 Sonnet Fusion Cases, and 4 internally mounted.

For quite some time, SMART Utility would give me SMART data on the 6 internal drives (4 of which are attached to a Sonnet card), and 4 of the 20 external drives, which would randomly show up.

Last week I updated SMART Utility to 3.12 and with that update, SMART data on all 24 drives is now visible.

Oct. 12, 2014

item.198103

Tracy Valleau

A couple of comments based on my experience:
SMART has various implementations: some include extended attributes; some do not. Some versions maintain logs; some do not. Failure threshold numbers vary from model to model. Implementation on the drive is often considered a "trade secret." Controllers and motherboards may pass along this, but not that.

If you're holding a failed drive in your hand, there is about a 50/50 chance that SMART could have predicted it. SMART cannot predict the failure of bearings nor a fried electronic component or crashed head.

Some third party software is smarter (sorry) about using the data than others may be. For example, some software will report a drive as failing if the log shows just one reallocated sector, even if that reallocation happened 4000 hours ago, and there's been no problem since (as is the case with one of my removable drives, and SoftRAID's monitor.)

Personally, I've chosen SMART Utility for my reporting needs.

However, one needs to remember that your drive's SMART is statistically only slightly "better than nothing" at predicting drive failure. In some cases, it's obvious, but others require judgement to analyze.

Finally, remember that drive manufacturers (all two of them these days, at least for spinners) are not stupid. They warranty their drives for about how long they will actually last. (They don't want to be replacing them with any frequency.) Usually a drive will last longer, but "usually" is on your dime, not theirs.

I run a combination of SSDs and spinners. My choice for SSDs are Samsung EVO and Crucial M100. I find the Hitachi 5-years a better deal than the WD Blacks, because the are faster. I buy 1-year drives only for storage. Finally, I replace my drives based on their number of hours of use, (and rotate them into removable storage-only use) and don't wait for one to be marked as "failing" by SMART.

YMMV.

Oct. 13, 2014

item.198123

David Anders

http://ifixit.com has detailed howto's on hard drive replacement in Mac Minis.

Oct. 14, 2014

item.198150

Richard Hart

I am happy to report that all SMART utilities work just fine over Thunderbolt. For example, SMART Utility running on a new Mac Pro can analyze any drives in an OWC Thunderbay 4. Still no eSATA solution, though.

Feb. 2, 2015

item.205364

Colleen Thompson

My old OWC On-THE-Go FW/USB2 2.5" enclosure supported SMART via the SAT-SMART kext, at least till I dropped it and broke it the other day. OWC Support tells me they don't have *any* currently available portable enclosures that support SMART. Can anyone confirm or, preferably, refute this?

Feb. 3, 2015

item.205393

Samuel Herschbein

Colleen Thompson asked if OWC's enclosures support the SAT-SMART kext.

I just checked two OWC Mercury On-The-Go Pro enclosures - both work with SAT-SMART and SMART Utility in Snow Leopard. Both were hooked up to my Mac via FireWire 800. The old one has USB 2, the new one has USB 3.

Jun. 11, 2015

item.211868

Alfred Cellier

I hope someone can shed some light, which MacInTouch is so good at.

We have an iOmega eGo FW800 enclosure, inside which is a Samsung HD204UI 2TB. This one seems to have been made in late 2011, after the early 2011 firmware update that avoided bad effects with the SatSMART kext.

Recently we ran SMART tests with several programs that present graphical "thermometer" results, and they made it appear that one attribute was in big trouble, even though saying "passed" overall.

This particular SMART attribute seemed worrisome:

ID# ATTRIBUTE NAME FLAG VALUE WORST THRESH TYPE UPDATED FAILED RAW VALUE
197 Current Pending Sector Count 0x0032 252 100 000 Old age Always Never 0

Assuming that it might indicate some bad sectors, we pulled the drive out and put it in several other, different enclosures, the latest being a JacobsParts SuperSpeed USB 3.0 SATA Hard Drive Docking Station, which is SatSMART-compatible. Always the SMART results were similar.

We repartitioned from 10 or so down to 2 volumes, formatted, then erased free space, and formatted again - figuring this would write over any bad sectors. A long SMART test showed no errors. Attributes remain the same.

I'm beginning to be suspicious of the "worst=100", being it is such a round number, as creating a false interpretation in these SMART programs. Maybe there is actually nothing wrong with this drive.

Does anyone else have a Samsung HD204UI 2TB, that they can report their 197 attribute?

P.S. I tried testdisk and the only oddity was that superblock boot sector and backup differed in several characters, same locations for both partitions. And, superblock boot sector could not copy to backup, nor vice versa.

Jun. 20, 2015

item.212331

Kurt L.

Alfred Cellier reported some problems with a Samsung hard drive back on June 11th, and he noted that the SMART data was all clear except for one attribute, Current Pending Sector Count.

We just experienced a hard drive failure in my wife's 2010 iMac, and the affected drive was a Samsung HD502HJ (500GB rotating media). In our case, the drive was clearly failing. Carbon Copy Cloner was reporting that certain files could not be read, and when we brought it into the local independent Apple dealer they ran an Apple diagnostic that blazoned "FAILED" across the screen.

Like Alfred, I had checked SMART data as soon as I suspected that something was failing. And like him, I didn't see any obvious red flags in the SMART data: no reallocated sectors. It did report that the Current Pending Sector Count was non-zero (and I observed that it seemed to be increasing slightly with time).

Different vendors implement SMART in different ways. My takeaway on this episode was that Samsung doesn't implement much in the way of SMART reporting.

For those interested in divining hard disk health based on SMART data, this web page reports on which attributes seem to matter:

Using SMART to accurately predict when a hard drive is about to die

Jun. 22, 2015

item.212345

Steven Gold

An interesting tip for checking SMART status: Disk Utility!

Sometime ago (I think it was Lion), Apple gave Disk Utility the ability to display the SMART values for internal drives:

Open Disk Utility and select the hard drive (not the volumes under it). Press the "Info" button on the toolbar and a window with the SMART values will open.

The values returned will vary based on the make and model of the unit, but, in general, look at the values with the word "Error", "Retries", "Reallocation", or "Uncorrectable" in the label. A drive functioning perfectly will have all zeros for the counts. Anything else *may* indicate a real or potential issue. A very small but non-zero value may indicate a temporary error. Power off the computer, turn it back on and see if the count is now zero.

item.212367

Mike Stupinski

I assume this must require Yosemite. I have Disk Utility Version 13 (517) under OS 10.9.5, but when I select the internal drive under DU I find no 'Info' button. At the bottom of the window 'S.M.A.R.T. Status' simply says "Verified."

Jun. 23, 2015

item.212373

Don Mac

For my internal 1 TB HDD, the SMART data shows numbers and letters in the data. For Example, for Raw Error Rate, its shows a number 000003F980CE... or Seek Error Rate, 0000006AA55A. a couple of the other readings have a combination of numbers and letters, also.

When I run a Smart Check using Tech Tools Pro, the HDD passes everything. So, I don't see how to actually read the data from the SMART information in Disk Utility.

When I look for SMART info on my self installed SATA internal SSD, it just shows SMART Verified. Same with the two SSDs connected through Thunderbolt.

item.212381

Mike Stupinski

Aha! I found the SMART data Steve Gold was talking about. It's 'Get Info' under the file menu, where it appears in a fair-size list of interesting information.

Guess I should keep looking before posting!

[I also searched unsuccessfully for a long time for SMART details in Disk Utility, then I realized that FileVault 2 / Core Storage formatting seems to hide SMART data in Disk Utility, though it's still available via SMART Utility. Very confusing! -Ric Ford]

item.212384

Richard Theriault

No, this doesn't require Yosemite. I'm running Mountain LIon (10.8.5) and tried it. The Info button was second from the left in my Disk Utility toolbar (traditional i in a blue circle) and a click on it produced the chart described. I was pleased to see zeros in the important places. It is nice to know about this!

item.212388

Colleen Thompson

I assume this must require Yosemite. I have Disk Utility Version 13 (517) under OS 10.9.5, but when I select the internal drive under DU I find no 'Info' button.

I have 10.9.5, and it's there. It's up in the toolbar at the top of the DU window. Easily overlooked, it's a blue circle with the white italicized i.

item.212401

Neal Tobochnik

For OS 9.5 there should be an info button in the Toolbar. If not go to Windows> Customize Toolbar and put the button onto the toolbar.

item.212406

Mike Stupinski

It's even *more* easily overlooked if it isn't there, Colleen! I had to customize my tool bar to find that button.

Jun. 24, 2015

item.212435

MacInTouch Reader

Richard Theriault noted:

No, this doesn't require Yosemite. I'm running Mountain LIon (10.8.5) and tried it. The Info button was second from the left in my Disk Utility toolbar (traditional i in a blue circle) and a click on it produced the chart described.

I'm on 10.8.5 and all I see in the Info sheet referencing SMART is "S.M.A.R.T Status: Verified".
Is there some Terminal command required to see the rest of the SMART statistics?

item.212410

Davide Guarisco

Many thanks to Steven Gold and others, who found a way to show SMART attributes in Disk Utility.

This is one of the most hidden and obscure features of OS X.

Here's a comparison of the output from Disk Utility's Get Info vs. what is returned by the smartctl tool on one of my hard drives.

Disk Utility:

S.M.A.R.T. Status : Verified
Raw Error Rate : 000000000000
Spinup Time : 000000000E4A
Start/Stop Count : 00000000007E
Reallocated Sectors : 000000000000
Seek Error Rate : 000000000000
Power-On Hours : 00000000358C
Spinup Retries : 000000000000
Calibration Retries : 000000000000
HDD Temperature : 00000000002F
Reallocated Sector Events : 000000000000
Current Pending Sectors : 000000000000
Offline Scan Uncorrectable Sectors : 000000000000
CRC Error Rate : 000000000000
Multi-Zone Error Rate : 000000000000

Smartctl:

ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
1 Raw_Read_Error_Rate 0x002f 200 200 051 Pre-fail Always - 0
3 Spin_Up_Time 0x0027 186 174 021 Pre-fail Always - 3658
4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 126
5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0
7 Seek_Error_Rate 0x002e 200 200 000 Old_age Always - 0
9 Power_On_Hours 0x0032 082 082 000 Old_age Always - 13708
10 Spin_Retry_Count 0x0032 100 100 000 Old_age Always - 0
11 Calibration_Retry_Count 0x0032 100 100 000 Old_age Always - 0
12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 107
192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 102
193 Load_Cycle_Count 0x0032 200 200 000 Old_age Always - 23
194 Temperature_Celsius 0x0022 100 091 000 Old_age Always - 47
196 Reallocated_Event_Count 0x0032 200 200 000 Old_age Always - 0
197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0030 200 200 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0
200 Multi_Zone_Error_Rate 0x0008 200 200 000 Old_age Offline - 0

So to answer Don Mac's question, the numbers shown in Disk Utility are the raw, hexadecimal values of the SMART attributes. For example, temperature is reported as 0x2F, which corresponds to 47, as reported by smartctl. Same for power-on hours, etc.

If you carefully compare the two lists of attributes, you will see that three of them are missing from Disk Utility's output: Power_Cycle count, Power-Off Retract_Count, and Load_Cycle_Count. Also, Disk Utility does not show the SMART ID. (I am running Mavericks' Disk Utility).

So you may still want to use a specialized tool to get the full SMART info, but the most important attributes to assess the health of a disk drive (Reallocated sector count, Pending sector reallocation, and drive temperature) are shown directly by Disk Utility.

item.212419

Peter Engels

Steven Gold says:

An interesting tip for checking SMART status: Disk Utility!

It's true that (as others have also reported), Disk Utility can provide SMART status info. However, when my hard disk failed, Disk Utility did not report on SMART status until it was too late -- the disk was unusable by the time a Disk Utility check reported a failing disk.

Fortunately, I was using Carbon Copy Cloner for a daily backup; it started reporting bad disk data three days before SMARTReporter, and a full week before Disk Utility, which I had been checking every day. No data was lost, thanks to CCC, which I highly recommend.

item.212420

Peter Neame

SMART data in Disk Utility? I seem to be able to access it by adding "Log" to the toolbar, then doing "getInfo" (which gives exceptionally little info, as Ric points out) followed by "log", whereupon the SMART data is the last entry. Very convoluted.

FileVault enabled and OS X 10.10.3

[With Core Storage enabled on my laptop's internal SSD, Disk Utility (OS X 10.9) shows only a Logical Volume Group in the left-hand pane - there is no physical device to select for Info, such as SMART data. This seems really perverse to me when it's simple and straightforward in SMART Utility by contrast. -Ric Ford]

item.212451

MacInTouch Reader

This is pretty confusing, and another twist is that (in Mavericks) SMART status is not supported for any drives other than the one built into the Mac. My startup drive is an external Firewire drive, and - like other external drives - SMART status is not supported for it.

[The SAT SMART driver, which can be installed with help from SMART Utility, does provide SMART support for some external FireWire and USB (including both USB 2 and 3) devices. -Ric Ford]

item.212448

Fred Brock

Sadly, all Disk Utility can tell me about SMART Status for my 480 GB Seagate SSD (ST480HM000 in a White MacBook early 2009) is the same message System Information reports... and other SMART utilities as well. "S.M.A.R.T. status: Not Supported." Thanks, Seagate!

Jun. 25, 2015

item.212456

Colleen Thompson

It's true that (as others have also reported), Disk Utility can provide SMART status info. However, when my hard disk failed, Disk Utility did not report on SMART status until it was too late -- the disk was unusable by the time a Disk Utility check reported a failing disk.

What people have been talking about recently is how to view individual SMART attributes in Disk Utility. If you wait for DU to warn you that a disk is failing, then it is, indeed, often too late. You don't say if you were checking the attributes in the case of your failed disk.

The vast majority of hard drive failures I see is due to bad blocks (sectors). The standard threshold for a bad blocks failure is pretty high, much higher than many people believe is reasonable. So even if it knows about some bad blocks, DU won't report a drive as failing. However, if you check the attributes and see that there are some bad blocks reported or pending, you can take steps.

In SMART Utility, the author devised his own warning thresholds for failing and failed disks. That's why it's so much more useful than DU.

Fortunately, I was using Carbon Copy Cloner for a daily backup; it started reporting bad disk data three days before SMARTReporter, and a full week before Disk Utility, which I had been checking every day. No data was lost, thanks to CCC, which I highly recommend.

CCC's log is an excellent indicator of bad blocks. If you see I/O errors, you know you have a disk problem. There's no threshold business involved, unless in this case you consider the threshold to be one file.

There are other utilities, of course, for checking SMART, like Tech Tool Pro. I'm most familiar with SMART Utility. No affiliation, just a satisfied customer, etc.

item.212459

David Weatherston

With reference to a MacInTouch reader who is looking for a command that will bring up the S.M.A.R.T. info shown in Davide Guarisco's post, I got it all by clicking the blue-dot-with-italic-i button on Disk Utility's toolbar in machines running 10.7.5 and 10.9.4. On the other hand, all I got from the blue-dot-with-italic-i button under 10.5.8 was "S.M.A.R.T. Status : Verified", so while S.M.A.R.T. info has been available for some time, it hasn't been there forever.

I'm baffled by the use of hex, though. While most users (including me) would have difficulty in interpreting the data if rendered in familiar numbers, throwing up hex seems unnecessarily user-hostile, even for present-day Apple.

[For what it's worth, SMART Utility (3.0.2) offers full SMART data reports on Mac OS X 10.4 (for a modern hard drive or SSD), though Disk Utility (10.5.6) doesn't, as you noted. -Ric Ford]

item.212487

Colleen Thompson

...throwing up hex seems unnecessarily user-hostile, even for present-day Apple.

Since Apple knows that only more sophisticated users will ever look at this feature, it's basically Apple thumbing its nose at us and saying "You think you want advanced features? Let's see how you like this!"

I think it's more spiteful than hostile.

item.212499

Stephen Hart

David Weatherston wrote:

"I'm baffled by the use of hex, though. While most users (including me) would have difficulty in interpreting the data if rendered in familiar numbers, throwing up hex seems unnecessarily user-hostile, even for present-day Apple."

Hex is what the drive reports. Utilities like Smart Utility interpret the hex.

BTW, Disk Utility does not report Smart status for my 3T Fusion drive (it's a logical volume), but Smart Utility does.

item.212515

David Weatherston

As bit of followup on SMART data and its interpretation - coincident to this discussion, I suddenly have an overheating hard drive (54 celsius). Disk Warrior tossed up a warning dialogue but SMART Utility is silent. Surprising - I would have thought heat is a major flag for problems (whether it is or not, I'm off to the repair depot tomorrow morning).

Not to quibble (well, all right), but perhaps my earlier post shouldn't have said, "I'm baffled by the use of hex..." but "I'm baffled by Apple's failure to have Disk Utility translate the computer's internal language." Didn't they start out saying that they made "Computers for the rest of us"?

item.212520

Bill DeFelice

Please forgive me if this was already posted, but I found this quite insightful. It reaffirmed me being leary of 3TB drives and broke the bad news about two of the more popular drive manufacturers.

When will your hard drive fail?
One storage provider compiles failure rates for major hard drive vendors and a pattern emerges.

Jun. 26, 2015

item.212548

David Charlap

Bill DeFelice wrote:

"Please forgive me if this was already posted, but I found this quite insightful. It reaffirmed me being leary of 3TB drives and broke the bad news about two of the more popular drive manufacturers."

This article is an extremely brief summary of a much bigger and ongoing report by Backblaze.

They have reported serious problems with Seagate's 3TB Barracuda dives. They did not report these problems with 3TB drives from other manufacturers and they have not reported problems with Seagate's 4TB drives.

I recommend you do some web searching to read up on Backblaze's most recent reports from their blog, which identifies specific makes and models that they found to be particularly good and bad. You'll find that every manufacturer has some good and some bad models, and that every size has some good and some bad models.

Also note that Backblaze only uses consumer-grade drives. They do not use enterprise-class drives (they explain why in some of their blog posts), so you won't find any information about any enterprise-class drive in their reports. Many people (myself included) find them more reliable than consumer drives.

Finally, you should note that the article you cited points out that the failure rates for Seagate 3TB Barracuda drives has been going down recently. The article states that we may have been seeing an anomaly from a bad manufacturing run, and not necessarily a problem with the design.

Next Page...


MacInTouch Amazon link...

Talk to MacInTouch     Support  •  Find/Go