Internal Drive Reports Failing SMART Tests

--On October 20, 2005 12:31:23 PM -0700 James White
<jamesrwhite2comcast.net> wrote:

> For what its worth, here's my recommendation:
>
> Step 1 - As cheap as they are and given the odds of success (see other
> posts), is it really worth trying to fix it? - BUY A NEW DRIVE! (now
> who said you can't throw money at a problem)

Seagate now offers a 5 year warranty on their new drives, and they don't
appear to be any more expensive than Western Digital or Maxtor at the
CompUSA here in town.

Kevin

Kevin van Haaren said:

>Seagate now offers a 5 year warranty on their new drives, and they don't
>appear to be any more expensive than Western Digital or Maxtor at the
>CompUSA here in town.


TigerDirect.com has a Seagate 250GB 7200RPM 8MB Buffer IDE Hard Drive for
$70 after $60 in rebates.

<http://www.tigerdirect.com/applications/SearchTools/item-details.asp?sourc
eid=qIZcQ1iAZ5TaAjz4hiHM&siteid=0037550214&CMP=AFC-AFFIL&Sku=THD-250A&EdpNo
=1265028&SRCCODE=BEFREE>

or

http://seagate250.notlong.com



Randy B. Singer
Co-Author of: The Macintosh Bible (4th, 5th and 6th editions)

Routine OS X Maintenance and Generic Troubleshooting
http://www.macattorney.com/ts.html

I'll echo everyone else here. When SMART says your disk is failing, RUN, DO NOT WALK, to make a full backup and install a new disk. As someone already said, disks are too cheap, and your data is too valuable, to screw around.

SMART measures about a dozen variables (varies by drive mfr). It doesn't just measure the number of bad blocks. It measures trends over time, like how many bad blocks are accumulating, whether the number of read and seek errors is increasing, and warns when they go over a threshold.

Call Maxtor or check their web site to see if your disk is still under warranty. Even if the iMac is out of warranty from Apple, the drive may still be under warranty from Maxtor. Both HP and IBM have replaced drives for me under warranty, even after the OEM warranty for the entire enclosure had already expired.

--
David

DiskWarrior has an automatic SMART status monitor, and so does the free Julian Mayer's SMART REPORTER

Does anyone know if these two SMART monitors are functionally equivalent? Or is there another, bettor SMART monitor I should look at?

Obviously, they have feature differences, but will they both report based on the exact same criteria? I've tried both of them, and both can send an email to my cellphone, at least the test email works ok. I (luckily) have not yet seen real disk failure since I recently installed these nifty S.M.A.R.T. monitors. So I can't comment on the reliability or accuracy.

Question: Will any/all SMART drive status monitor(s) tap into the same exact pass/fail self test of the drive hardware? Some people have reported varying results. This is such an important function, and we need to know how reliable the status monitors are. Can anyone speak to this?
Dave
MacMedix

One of my hard drives once reported a failing status in Disk Utility. Needless to say, I backed all data up. The interesting thing is that since then the SMART diagnostics for that same disc has been reporting an OK status for many weeks now. How can this be explained? I decided to keep that hard drive, but I only use it for non essential duties (game installation for example).

Another question: none of my external firewire cases report the SMART status of the drive inside. Disk Utility reports "not supported". Does anybody know of any firewire case that does report SMART status?

TIA

Jean-Denis

Thanks to Paul, Travis, Chris, Jason, Sander, Bill, John, James, Randy,
and David for their advice. The general consensus seems to be that
failing the S.M.A.R.T. status means more than just a bad block failure
and probably indicates a continuing and worsening problem, therefore
back up the drive and replace it as soon as possible. Chris and David,
my drive's warranty expired in March of this year (thanks for the
reminder to check-I had assumed that since my computer's warranty
period was over that the drive's was also). James, SMARTReporter does
indeed report the drive as failing and recommends I replace it. Kevin
and Randy, thanks for the tips on Seagate drives at CompUSA and
TigerDirect.

Next question. I was told that I would need a 5400RPM drive for this
particular computer. I'm using a 7200RPM external Firewire drive as my
boot drive right now, but since this is internal, I will need to get a
5400RPM instead. Is this true? Also, what is the maximum capacity I can
have for my internal drive on this computer. Is there anything else I
need to careful of when buying a new drive for this computer? Just to
remind you, I am using a 800MHz Flat-Panel iMac G4.

-Ray Still
   tregarth (at) fastmail (dot) fm

>Does anybody know of any firewire case that does report SMART status?

The Granite Digital FireWire Drive Bays do...

<http://www.granitedigital.com/>

cheers... -Adam

On 10/26/05 at 1:10 PM, tregarthfastmail.fm (Ray Still) wrote:

> Next question. I was told that I would need a 5400RPM drive for this
> particular computer. I'm using a 7200RPM external Firewire drive as
> my boot drive right now, but since this is internal, I will need to
> get a 5400RPM instead. Is this true? Also, what is the maximum
> capacity I can have for my internal drive on this computer. Is there
> anything else I need to careful of when buying a new drive for this
> computer? Just to remind you, I am using a 800MHz Flat-Panel iMac G4.

The only reason I can think of off the top of my head is temperature; a
7200 RPM drive might run somewhat hotter than a 5400 RPM drive just
because it spins faster, though I would read the drive spec sheets for
definitive information; one company's 7200 RPM drive might run cooler
than another company's 5400 RPM drive.

Unfortunately, the copy of the iMac G4 developer's tech note that I have
(from 2002) doesn't list any internal temperature specs that I could
find, so I can't tell for sure what temperature you should shoot for
with a replacement drive; all I have is a vague memory of the iMac G4's
cooling being carefully planned to fit in the small case with just the
one ventilation fan.

The iMac G4 has an ATA/66 bus, according to the tech note; that might
have been why you'd been told to get a slower drive, but IIRC that
shouldn't matter as the drive should downshift to match a slower bus.

Travis Butler
tbutlermac.com

At 13:10 -0700 UTC, on 2005/10/26, MacMedix wrote:

> <http://www.alsoft.com/DiskWarrior/>DiskWarrior has an automatic SMART
>status monitor, and so does the free <http://homepage.mac.com/julianmayer/>
>Julian Mayer's SMART REPORTER
>
> Does anyone know if these two SMART monitors are functionally equivalent?

My impression is that the only magic to these utilities is that they retrieve
the disk's SMART status periodically whereas with Disk Utility you need to
remind yourself to do so now and then.

You can easily automate it yourself by having some script periodically call
/usr/sbin/diskutil. I did so with AppleScript. See
<http://www.euronet.nl/~tekelenb/software/>. You can't miss it - it's called
SMARTer :) Free. Code is editable, so you can see how it works and change it
to whatever you like.

[...]

>This is such an important function, and we need to know how
>reliable the status monitors are.

I think the evidence we've seen in this thread shows that it is not reliable
- not only may a "Failing" disk keep on working, a "Verified" disk may be
broken. I think we can consider SMART another useful indicator tool, nothing
more. The real magic is in good backups.


--
Sander Tekelenburg, <http://www.euronet.nl/~tekelenb/>

On 10/26/05 at 1:10 PM, tregarthfastmail.fm (Ray Still) wrote:

>Next question. I was told that I would need a 5400RPM drive for
>this particular computer. I'm using a 7200RPM external Firewire
>drive as my boot drive right now, but since this is internal, I
>will need to get a 5400RPM instead. Is this true? Also, what is the
>maximum capacity I can have for my internal drive on this computer.
>Is there anything else I need to careful of when buying a new drive
>for this computer?

Assuming a constant power form, the spin rate for a drive will not matter. You will get better performance with a higher spin rate.

By power form, I mean the voltage and current consumption for the drive. These are the factors that are important.

The other factor that is important is the physical size of the drive. And given

>I am using a 800MHz Flat-Panel iMac G4.

the only significant factor (other than performance) will be the physical size of the drive.

To give you a specific example, when I replaced the hard drive on my old PowerWave machine running OS 9 I went from a 4 GB (either 5400 or 7200 rpm, I don't specicially recall which) to a 18 GB 15,000 rpm drive (Seagate Cheetah). It worked fine and is still working fine.

I have a drive on my iMac FP where TechTool is reporting a SMART Failure (speifically, Attribute 4: Start/Stop Count). However, neither DU, DiskWarrior, nor SMARTReporter are listing any errors on this drive?

There *was* clearly something wrong previously. DU was unable to do a Repair on the drive, but DiskWarrior was. Still, after the repair, only TechTool reports the SMART Failure.

I am running Mac OS 10.4.5 and have latest versions of DiskWarrior, techTool, and SMARTReporter.

Ideas?

Thanks.

Chuck Coulson

On 3/6/06 at 10:30 AM, chuck.coulsoncomcast.net (Chuck Coulson)
wrote:

>I have a drive on my iMac FP where TechTool is reporting a SMART
>Failure (speifically, Attribute 4: Start/Stop Count). However,
>neither DU, DiskWarrior, nor SMARTReporter are listing any errors
>on this drive?

>Ideas?

Drives are cheap. Replacing any significant data can be extremely painful to impossible. Unless you really don't care about the data, replace the drive.

Le 06-03-08 à 02:03, Bill Rowe a écrit :

>> I have a drive on my iMac FP where TechTool is reporting a SMART
>> Failure (speifically, Attribute 4: Start/Stop Count). However,
>> neither DU, DiskWarrior, nor SMARTReporter are listing any errors
>> on this drive?
>
> Drives are cheap. Replacing any significant data can be extremely
> painful to impossible. Unless you really don't care about the data,
> replace the drive.

Totally agreed. SMART isn't a warning that the drive is in the
process of failing - it's a warning that the drive is *going* to
fail. I would probably take this under consideration and buy a new
drive ASAP.

Neil

At 19:02 -0800 UTC, on 2006-03-09, Neil Lee wrote:

> Le 06-03-08 à 02:03, Bill Rowe a écrit :
>
>>> I have a drive on my iMac FP where TechTool is reporting a SMART
>>> Failure (speifically, Attribute 4: Start/Stop Count). However,
>>> neither DU, DiskWarrior, nor SMARTReporter are listing any errors
>>> on this drive?
>>
>> Drives are cheap. Replacing any significant data can be extremely
>> painful to impossible. Unless you really don't care about the data,
>> replace the drive.
>
> Totally agreed. SMART isn't a warning that the drive is in the
> process of failing - it's a warning that the drive is *going* to
> fail. I would probably take this under consideration and buy a new
> drive ASAP.

Agreed. But that ignores what I took as the OP's real question: why one app
would report SMART failure and another does not. My impression was that it is
the drive itself that provides the SMART status. Unless that's not the case,
every app that reports a drive's SMART status should therefore be reporting
the same. If it doesn't, the problem might be with the software that queries
the drive for its SMART status.


--
Sander Tekelenburg, <http://www.euronet.nl/~tekelenb/>

On 3/10/06 at 2:24 PM, tekelenbeuronet.nl (Sander Tekelenburg)
wrote:

>Agreed. But that ignores what I took as the OP's real question: why
>one app would report SMART failure and another does not. My
>impression was that it is the drive itself that provides the SMART
>status. Unless that's not the case, every app that reports a
>drive's SMART status should therefore be reporting the same. If it
>doesn't, the problem might be with the software that queries the
>drive for its SMART status.

The OP also indicated he was having problems with the drive *before* he got the result of TechTool showing a failing SMART status and the other programs reporting a passed SMART status. Given that, I am would not be inclined to believe the issue is faulty reporting by TechTool. Instead, I would *guess* DU and DW simply don't report certain SMART errors for whatever reason. And if it were my drive, I would be getting a replacement drive.

The issue of why different programs give different SMAET status for the drive is essentially undecidable for a typical user. That is lacking access to source code for the programs there is no way to know for certain why a given program reports SMART status differently than another. And lacking a "gold" standard to determine the SMART status of a drive, there is no way to determine which of several conflicting reports is correct.

On Mar 10, 2006, at 2:24 PM, Sander Tekelenburg wrote:

> At 19:02 -0800 UTC, on 2006-03-09, Neil Lee wrote:
>
>> Le 06-03-08 à 02:03, Bill Rowe a écrit :
>>
>>>> I have a drive on my iMac FP where TechTool is reporting a SMART
>>>> Failure (speifically, Attribute 4: Start/Stop Count). However,
>>>> neither DU, DiskWarrior, nor SMARTReporter are listing any errors
>>>> on this drive?
>>>
>>> Drives are cheap. Replacing any significant data can be extremely
>>> painful to impossible. Unless you really don't care about the data,
>>> replace the drive.
>>
>> Totally agreed. SMART isn't a warning that the drive is in the
>> process of failing - it's a warning that the drive is *going* to
>> fail. I would probably take this under consideration and buy a new
>> drive ASAP.
>
> Agreed. But that ignores what I took as the OP's real question: why
> one app
> would report SMART failure and another does not. My impression was
> that it is
> the drive itself that provides the SMART status. Unless that's not
> the case,
> every app that reports a drive's SMART status should therefore be
> reporting
> the same. If it doesn't, the problem might be with the software
> that queries
> the drive for its SMART status.
>

The S.M.A.R.T. reporting by the various drives is not created equal.
Different drives report different collections of information.

Not all people who write utilities that read S.M.A.R.T. are equal
this long after their creation: some are better programmers than
others; some have access to more samples than others, and so on.

Remember, there are two kinds of disk drives: those that have
failed, and those that haven't failed yet. S.M.A.R.T. is a worthy
attempt to provide a signal when a drive is preparing to move from
the second group to the first.

It can also be good for the customer, as many manufacturers accept a
suitable S.M.A.R.T. report as a basis for no- (or few-) questions
asked replacements, before the final move to the "drives that have
failed" group.

   --John

--On March 10, 2006 2:24:24 PM -0800 Sander Tekelenburg
<tekelenbeuronet.nl> wrote:

> Agreed. But that ignores what I took as the OP's real question: why one
> app would report SMART failure and another does not. My impression was
> that it is the drive itself that provides the SMART status. Unless that's
> not the case, every app that reports a drive's SMART status should
> therefore be reporting the same. If it doesn't, the problem might be with
> the software that queries the drive for its SMART status.

According to Seagate:
<http://www.seagate.com/support/kb/disc/smart.html>

"In an ATA/IDE environment, software on the host interprets the alarm
signal from the drive generated by the "report status" command of
S.M.A.R.T." and "The host system can evaluate the attributes and alarms
reported, in addition to the "report status" command from the disc."

Apparently the SCSI version operates more like you're describing:
"In a SCSI environment, the failure decision occurs at the disc drive, and
the host notifies the user for action."


I would guess that the one app might be overly sensitive to certain tests
or might be checking more SMART statuses than the other apps. When I run
TechTool Pro, the report shows each test it checked. You might see if you
can find out exactly what it is reporting as failing:

> S.M.A.R.T. Self-Checks
>
> Attribute Normal Worst Threshold Status
>
> 1 Read Raw Error Rate
> 51 48 34 Okay
> 3 Spin Up Time
> 95 95 0 Okay
> 4 Start/Stop Count
> 100 100 20 Okay
> 5 Reallocated Sectors
> 100 100 36 Okay
> 7 Seek Error Rate
> 76 60 30 Okay
> 9 Power On Hours
> 98 98 0 Okay
> 10 Spin Retry Count
> 100 100 34 Okay
> 12 Power Cycle Count
> 100 100 20 Okay
> 192 Power Off Retract Count
> 100 100 0 Okay
> 193 Load/Unload Cycle Count
> 38 38 0 Okay
> 194 Temperature
> 44 48 0 Okay
> 195 HW ECC Recovered
> 51 48 0 Okay
> 197 Current Pending Sector Count
> 100 100 0 Okay
> 198 Off-Line Scan Uncorrectable Sector Count
> 100 100 0 Okay
> 199 Ultra DMA CRC Error Count (Rate)
> 200 200 0 Okay
> 200 Write Error Count
> 100 253 0 Okay
> 202 Unknown
> 100 253 0 Okay
>
> S.M.A.R.T. Self-Checks <Passed>

Sander Tekelenburg wrote:
> Agreed. But that ignores what I took as the OP's real question: why one app
> would report SMART failure and another does not.

There are two types of SMART failures: Old age and prefail. The former
just means the drive is old (which is what start/stop count sounds
like); the latter supposedly means "drive will be dead within 24h".

Can S.M.A.R.T. Status change? You be the judge:

My iMac G5 (no iSight) was custom ordered with an Hitachi hds724040klsa80 400GB SATA drive. Last week, the machine started slowing down, inexplicably spinning the pizza for minutes at a time. The hardware verification on the Install DVD said there was no problem, yet it took 2 hours to do the disk check portion of the test!

Running Disk Utility also showed no problem. Alsoft's Disk Warrior showed no problem. Next day, same trouble. After waiting for 20 minutes, I forcibly rebooted from the Install DVD and ran Disk Utility.

S.M.A.R.T. Status: Failing This drive has reported a fatal hardware error to Disk Utility. If the drive has not failed completely, back up as much data as you can and then replace it with a working drive.

I called Apple. They told me this error might not be what it seems and that something else might be causing it. I would have to bring the machine in to a repair depot. They would not send a replacement drive part because it is not user-installable (contrary to their website http://docs.info.apple.com/article.html?artnum=300205 ).

Not one for giving up instantly, and not wanting to lose my data, I borrowed another iMac (from a friend on vacation) with enough free space to hold my data (210GB). I restarted the sick machine in FireWire Target Disk mode then connected the two together. Over the next two days, Super Duper failed twice to copy the data (and the process was painfully slow).

I was about to resort to GNU ddrescue but I was tired. I shut down everything and left it overnight.

The next morning I disconnected the ailing machine, disabled Spotlight Indexing on the working iMac, rebooted and ran Disk Utility, restarted the sick machine in FireWire Target Disk mode and then connected the two together.

In Disk Utility, I immediately unmounted the ailing hard drive. From the command line, I remounted the drive as READ ONLY. Super Duper made a complete copy to the rescue partition without complaint at what appeared to be normal speed.

I edited the /etc/hostconfig file to disable Spotlight indexing.

Next steps: disconnect, restart working machine in Target mode, reconnect, reboot ailing machine with Option key down and select the copy made by Super Duper.

I ejected the "internal ailing disk" then (and this will sound silly), I immediately went to iTunes and deauthorized the machine.

Everything was working fine. No files were missing.

Running Disk Utility and checking the internal drive yielded:

S.M.A.R.T. Status: Verified

I left the disk running all weekend. Still no errors on Monday morning.

I erased the ailing disk with a 7-pass security option to see if anything would happen. Everything was fine this morning.

I've installed a base OS to see if the machine will be happy today.

Here's what I've learned so far:

S.M.A.R.T. Status can change from Failing to Verified

disable Spotlight indexing immediately on any ailing machine

disable Spotlight indexing on any machine used for recovery

an overnight cooldown may be in order

after 90 days, Apple service becomes a nightmare even though hardware is warranted for 1 year

What can I do now? I have screen shots of the Disk Utility and ASP showing the failure status; however, the disk now reports itself as verified!!!

How come S.M.A.R.T. status isn't "permanent." Does anyone know?

Is it true that a S.M.A.R.T. Failure can be generated by something other than the disk (as claimed by the non-helpful people at Apple)?

Hi James:

Enjoyed your tale of woe and recovery re SMART. It was very illuminating. My only comment is that my father-in-law has an iMAC G5 1.8 that has been thru 2 motherboards and HD’s. One of the poorer machines Apple has engineered since the Performa. Despite all the brickabats thrown the way of Norton Utilities, it was always my HD utility of last resort until OS X. Tech Tool Pro and Drive Genius seem to work best at disk recovery up to 10.39 but Ihaven’t had enough experiencewithTiger to comment. Best of luck!
--
James Connolly
Oakland, CA 94611
james.connollysbcglobal.net