SATA drive problem, or other

Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS I)
would lock up, and it made sense to me in that I set a scratch disk on
the new internal drive, and I thought that perhaps the drive had spun
down, and that's why PS locked up when it tried to access it.

However, over the past few weeks, the lockups have been increasing, esp.
ironic in that I've been freeing up disk space on all drives.

Any diagnostics or cures you can think of?

--
John McWilliams

On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:

> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS I)
> would lock up, and it made sense to me in that I set a scratch disk on
> the new internal drive, and I thought that perhaps the drive had spun
> down, and that's why PS locked up when it tried to access it.
>
> However, over the past few weeks, the lockups have been increasing,
> esp. ironic in that I've been freeing up disk space on all drives.
>
> Any diagnostics or cures you can think of?

The advisable diagnostic measures you should take really depend on what
exactly you mean by "lock up".

--
JR

Jolly Roger wrote:
> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>
>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS
>> I) would lock up, and it made sense to me in that I set a scratch disk
>> on the new internal drive, and I thought that perhaps the drive had
>> spun down, and that's why PS locked up when it tried to access it.
>>
>> However, over the past few weeks, the lockups have been increasing,
>> esp. ironic in that I've been freeing up disk space on all drives.
>>
>> Any diagnostics or cures you can think of?
>
> The advisable diagnostic measures you should take really depend on what
> exactly you mean by "lock up".

Yeah, I coulda been a bit clearer.... The spinning beach ball, no
response to force quit, no Cmd-tab to another app.

--
John McWilliams

In article <Q6CdnRHxgZKSXRXYnZ2dnUVZ_uTinZ2d@comcast.com>,
John McWilliams <jpmcw@comcast.net> wrote:

> Jolly Roger wrote:
> > On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
> >
> >> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
> >> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS
> >> I) would lock up, and it made sense to me in that I set a scratch disk
> >> on the new internal drive, and I thought that perhaps the drive had
> >> spun down, and that's why PS locked up when it tried to access it.
> >>
> >> However, over the past few weeks, the lockups have been increasing,
> >> esp. ironic in that I've been freeing up disk space on all drives.
> >>
> >> Any diagnostics or cures you can think of?
> >
> > The advisable diagnostic measures you should take really depend on what
> > exactly you mean by "lock up".
>
> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
> response to force quit, no Cmd-tab to another app.

I've seen this behavior from OS X when a drive is going bad, yes. Is the
drive making odd noises? Have you tried switching to another disk for
scratch space, and maybe even unmounting the new drive, to see if the
problems go away?

--
"Those who enter the country illegally violate the law."
-- George W. Bush in Tucson, Ariz., Nov. 28, 2005

On 2006-12-19 22:47:39 -0600, John McWilliams <jpmcw@comcast.net> said:

> Jolly Roger wrote:
>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>>
>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS I)
>>> would lock up, and it made sense to me in that I set a scratch disk on
>>> the new internal drive, and I thought that perhaps the drive had spun
>>> down, and that's why PS locked up when it tried to access it.
>>>
>>> However, over the past few weeks, the lockups have been increasing,
>>> esp. ironic in that I've been freeing up disk space on all drives.
>>>
>>> Any diagnostics or cures you can think of?
>>
>> The advisable diagnostic measures you should take really depend on what
>> exactly you mean by "lock up".
>
> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
> response to force quit, no Cmd-tab to another app.

If this were a drive problem, you would most likely see corresponding
/var/log/system.log entries, such as "disk I/O error" and the like. Do
you see anything like that in the system log immediately after the
problem?

--
JR

ZnU wrote:
> In article <Q6CdnRHxgZKSXRXYnZ2dnUVZ_uTinZ2d@comcast.com>,
> John McWilliams <jpmcw@comcast.net> wrote:
>
>> Jolly Roger wrote:
>>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>>>
>>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>>>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS
>>>> I) would lock up, and it made sense to me in that I set a scratch disk
>>>> on the new internal drive, and I thought that perhaps the drive had
>>>> spun down, and that's why PS locked up when it tried to access it.
>>>>
>>>> However, over the past few weeks, the lockups have been increasing,
>>>> esp. ironic in that I've been freeing up disk space on all drives.
>>>>
>>>> Any diagnostics or cures you can think of?
>>> The advisable diagnostic measures you should take really depend on what
>>> exactly you mean by "lock up".
>> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
>> response to force quit, no Cmd-tab to another app.
>
> I've seen this behavior from OS X when a drive is going bad, yes. Is the
> drive making odd noises? Have you tried switching to another disk for
> scratch space, and maybe even unmounting the new drive, to see if the
> problems go away?

No noises at all, or any symptoms I can detect other than the lockups,
which conceivably aren't related to the additional HDD.

I've done stuff other than unmounting the drive, which, until I can get
a larger backup, needs to be in place most of the time. But I think I
can move a few things so I can keep it unmounted most of the time,
reversing what I just described as my normal flow.

--
John McWilliams

Jolly Roger wrote:
> On 2006-12-19 22:47:39 -0600, John McWilliams <jpmcw@comcast.net> said:
>
>> Jolly Roger wrote:
>>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>>>
>>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>>>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS
>>>> I) would lock up, and it made sense to me in that I set a scratch
>>>> disk on the new internal drive, and I thought that perhaps the drive
>>>> had spun down, and that's why PS locked up when it tried to access it.
>>>>
>>>> However, over the past few weeks, the lockups have been increasing,
>>>> esp. ironic in that I've been freeing up disk space on all drives.
>>>>
>>>> Any diagnostics or cures you can think of?
>>>
>>> The advisable diagnostic measures you should take really depend on
>>> what exactly you mean by "lock up".
>>
>> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
>> response to force quit, no Cmd-tab to another app.
>
> If this were a drive problem, you would most likely see corresponding
> /var/log/system.log entries, such as "disk I/O error" and the like. Do
> you see anything like that in the system log immediately after the problem?

If I were in the habit of looking at the logs, perhaps! <s>. I will make
myself do that soon....and D'Oh! on me.

In thinking about it I had been unconsciously assuming the logs would
say that, begging the question: what now?

--
John McWilliams

On 2006-12-20 09:40:45 -0600, John McWilliams <jpmcw@comcast.net> said:

> Jolly Roger wrote:
>> On 2006-12-19 22:47:39 -0600, John McWilliams <jpmcw@comcast.net> said:
>>
>>> Jolly Roger wrote:
>>>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>>>>
>>>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>>>>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS I)
>>>>> would lock up, and it made sense to me in that I set a scratch disk on
>>>>> the new internal drive, and I thought that perhaps the drive had spun
>>>>> down, and that's why PS locked up when it tried to access it.
>>>>>
>>>>> However, over the past few weeks, the lockups have been increasing,
>>>>> esp. ironic in that I've been freeing up disk space on all drives.
>>>>>
>>>>> Any diagnostics or cures you can think of?
>>>>
>>>> The advisable diagnostic measures you should take really depend on what
>>>> exactly you mean by "lock up".
>>>
>>> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
>>> response to force quit, no Cmd-tab to another app.
>>
>> If this were a drive problem, you would most likely see corresponding
>> /var/log/system.log entries, such as "disk I/O error" and the like. Do
>> you see anything like that in the system log immediately after the
>> problem?
>
> If I were in the habit of looking at the logs, perhaps! <s>. I will
> make myself do that soon....and D'Oh! on me.
>
> In thinking about it I had been unconsciously assuming the logs would
> say that, begging the question: what now?

We'll cross that bridge when we come to it. First, duplicate the
problem and let's see what, if anything, is logged.

How often does this problem occur, BTW?

--
JR

Jolly Roger wrote:
> On 2006-12-19 22:47:39 -0600, John McWilliams <jpmcw@comcast.net> said:
>
>> Jolly Roger wrote:
>>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>>>
>>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>>>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS
>>>> I) would lock up, and it made sense to me in that I set a scratch
>>>> disk on the new internal drive, and I thought that perhaps the drive
>>>> had spun down, and that's why PS locked up when it tried to access it.
>>>>
>>>> However, over the past few weeks, the lockups have been increasing,
>>>> esp. ironic in that I've been freeing up disk space on all drives.
>>>>
>>>> Any diagnostics or cures you can think of?
>>>
>>> The advisable diagnostic measures you should take really depend on
>>> what exactly you mean by "lock up".
>>
>> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
>> response to force quit, no Cmd-tab to another app.
>
> If this were a drive problem, you would most likely see corresponding
> /var/log/system.log entries, such as "disk I/O error" and the like. Do
> you see anything like that in the system log immediately after the problem?

Here are the relevant (I hope) entries for then:

Dec 19 18:47:54 jpmG5 kernel[0]: disk0s3: I/O timeout.
Dec 19 18:52:34 jpmG5 kernel[0]: disk0s3: I/O error.

There are about 8 repeats of the second line subsequent to the later time.

--
John McWilliams

On 2006-12-20 13:13:44 -0600, John McWilliams <jpmcw@comcast.net> said:

> Jolly Roger wrote:
>> On 2006-12-19 22:47:39 -0600, John McWilliams <jpmcw@comcast.net> said:
>>
>>> Jolly Roger wrote:
>>>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net> said:
>>>>
>>>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs RAM;
>>>>> IIRC, a Maxtor 160G. Occasionally, it would seem that Photoshop (CS I)
>>>>> would lock up, and it made sense to me in that I set a scratch disk on
>>>>> the new internal drive, and I thought that perhaps the drive had spun
>>>>> down, and that's why PS locked up when it tried to access it.
>>>>>
>>>>> However, over the past few weeks, the lockups have been increasing,
>>>>> esp. ironic in that I've been freeing up disk space on all drives.
>>>>>
>>>>> Any diagnostics or cures you can think of?
>>>>
>>>> The advisable diagnostic measures you should take really depend on what
>>>> exactly you mean by "lock up".
>>>
>>> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
>>> response to force quit, no Cmd-tab to another app.
>>
>> If this were a drive problem, you would most likely see corresponding
>> /var/log/system.log entries, such as "disk I/O error" and the like. Do
>> you see anything like that in the system log immediately after the
>> problem?
>
> Here are the relevant (I hope) entries for then:
>
> Dec 19 18:47:54 jpmG5 kernel[0]: disk0s3: I/O timeout.
> Dec 19 18:52:34 jpmG5 kernel[0]: disk0s3: I/O error.
>
> There are about 8 repeats of the second line subsequent to the later time.

Well there you go. So you have a hardware problem of some sort. Some
possible causes come to mind:

One possibility is the hard disk platter(s) have bad blocks - blocks
that have gone bad since the initial formatting. The solution for that
one is easy: perform a low level format of the drive to map out the bad
blocks automatically. Hopefully, the problem will be gone.

Another possibility is that the ATA controller to which the drive is
connected is faulty. This one is easy enough to rule out: swap both
hard drives and see if the problem persists (the "disk0s3" portion of
the error message in system.log will change, I believe - but the error
will be the same otherwise).

Ruling those two out, another possibility is that some part of the
drive hardware is faulty - the drive motor, read/write heads, circuit
board, or firmware is faulty. If you've gotten this far, I recommend
you run Disk Utility to see the SMART status of the drive. If the SMRAT
status is verified, then run TechTool Pro (or Deluxe, if you have
AppleCare) and see if it can tell you any more about the drive.
Ultimately, if you've gotten this far, you're looking at replacing the
drive. Is the drive under warranty?

--
JR

Jolly Roger wrote:
> On 2006-12-20 13:13:44 -0600, John McWilliams <jpmcw@comcast.net> said:
>
>> Jolly Roger wrote:
>>> On 2006-12-19 22:47:39 -0600, John McWilliams <jpmcw@comcast.net> said:
>>>
>>>> Jolly Roger wrote:
>>>>> On 2006-12-19 21:20:56 -0600, John McWilliams <jpmcw@comcast.net>
>>>>> said:
>>>>>
>>>>>> Installed a second SATA drive on my G5 dual @ 1.8 Meg, 1.25Gigs
>>>>>> RAM; IIRC, a Maxtor 160G. Occasionally, it would seem that
>>>>>> Photoshop (CS I) would lock up, and it made sense to me in that I
>>>>>> set a scratch disk on the new internal drive, and I thought that
>>>>>> perhaps the drive had spun down, and that's why PS locked up when
>>>>>> it tried to access it.
>>>>>>
>>>>>> However, over the past few weeks, the lockups have been
>>>>>> increasing, esp. ironic in that I've been freeing up disk space on
>>>>>> all drives.
>>>>>>
>>>>>> Any diagnostics or cures you can think of?
>>>>>
>>>>> The advisable diagnostic measures you should take really depend on
>>>>> what exactly you mean by "lock up".
>>>>
>>>> Yeah, I coulda been a bit clearer.... The spinning beach ball, no
>>>> response to force quit, no Cmd-tab to another app.
>>>
>>> If this were a drive problem, you would most likely see corresponding
>>> /var/log/system.log entries, such as "disk I/O error" and the like.
>>> Do you see anything like that in the system log immediately after the
>>> problem?
>>
>> Here are the relevant (I hope) entries for then:
>>
>> Dec 19 18:47:54 jpmG5 kernel[0]: disk0s3: I/O timeout.
>> Dec 19 18:52:34 jpmG5 kernel[0]: disk0s3: I/O error.
>>
>> There are about 8 repeats of the second line subsequent to the later
>> time.
>
> Well there you go. So you have a hardware problem of some sort. Some
> possible causes come to mind:
>
> One possibility is the hard disk platter(s) have bad blocks - blocks
> that have gone bad since the initial formatting. The solution for that
> one is easy: perform a low level format of the drive to map out the bad
> blocks automatically. Hopefully, the problem will be gone.
>
> Another possibility is that the ATA controller to which the drive is
> connected is faulty. This one is easy enough to rule out: swap both hard
> drives and see if the problem persists (the "disk0s3" portion of the
> error message in system.log will change, I believe - but the error will
> be the same otherwise).
>
> Ruling those two out, another possibility is that some part of the drive
> hardware is faulty - the drive motor, read/write heads, circuit board,
> or firmware is faulty. If you've gotten this far, I recommend you run
> Disk Utility to see the SMART status of the drive. If the SMRAT status
> is verified, then run TechTool Pro (or Deluxe, if you have AppleCare)
> and see if it can tell you any more about the drive. Ultimately, if
> you've gotten this far, you're looking at replacing the drive. Is the
> drive under warranty?

Drive is over a year old.... but the SMART readings from Disk Utility
say 'verified' for both the drives. TechTool Pro is running in the
background now, and is half way through. I had already run it on
everything but the surface scan.

Since I have been thinking about a new drive anyway, putting in a new
one might be simplest, and thus backing up and reformatting the
potentially dodgey drive would be easiest. Swapping the drives might be
doable on the G5, but do I have to change any settings on the drives? Or
does the one in the "A" Bay (where the original drive sits)
automatically become Master? Or was that done away with with Serial ATA
drives. I can't recall if I set any pins on that second drive I put in
rather hurriedly.

--
John McWilliams

In article <W5GdnakA7pryNBTYnZ2dnUVZ_qisnZ2d@comcast.com>,
John McWilliams <jpmcw@comcast.net> wrote:
>
>Drive is over a year old.... but the SMART readings from Disk Utility
>say 'verified' for both the drives. TechTool Pro is running in the
>background now, and is half way through. I had already run it on
>everything but the surface scan.

If you have a more detailed SMART utility, check the UDMA CRC error
count. I had a drive (brand new, a Western Digital Caviar) which
didn't fail SMART but kept getting DMA CRC errors; I assume either a
bad or incompatible controller. Symptoms were the same -- timeouts
and failures.
--
There's no such thing as a free lunch, but certain accounting practices can
result in a fully-depreciated one.

Matthew T. Russotto wrote:
> In article <W5GdnakA7pryNBTYnZ2dnUVZ_qisnZ2d@comcast.com>,
> John McWilliams <jpmcw@comcast.net> wrote:
>> Drive is over a year old.... but the SMART readings from Disk Utility
>> say 'verified' for both the drives. TechTool Pro is running in the
>> background now, and is half way through. I had already run it on
>> everything but the surface scan.
>
> If you have a more detailed SMART utility, check the UDMA CRC error
> count. I had a drive (brand new, a Western Digital Caviar) which
> didn't fail SMART but kept getting DMA CRC errors; I assume either a
> bad or incompatible controller. Symptoms were the same -- timeouts
> and failures.

The only other SMART utility I could find readily at Versiontracker is
called SMARTReporter gives a little report such as

"2006-12-20 15:34:44.399 SMARTReporter[648] ATA Drive:
'ST3160827AS' - SMART condition not exceeded, drive OK!"

for each of my 4 partitions, two on each drive. Is the controller part
of the HD or part of the G5?

And my Techtool Surface scan reported no bad blocks on either partition
on the possibly goofy drive.

--
John McWilliams

On 2006-12-20 15:21:51 -0600, John McWilliams <jpmcw@comcast.net> said:

>>> Here are the relevant (I hope) entries for then:
>>>
>>> Dec 19 18:47:54 jpmG5 kernel[0]: disk0s3: I/O timeout.
>>> Dec 19 18:52:34 jpmG5 kernel[0]: disk0s3: I/O error.
>>>
>>> There are about 8 repeats of the second line subsequent to the later time.
>>
>> Well there you go. So you have a hardware problem of some sort. Some
>> possible causes come to mind:
>>
>> One possibility is the hard disk platter(s) have bad blocks - blocks
>> that have gone bad since the initial formatting. The solution for that
>> one is easy: perform a low level format of the drive to map out the bad
>> blocks automatically. Hopefully, the problem will be gone.
>>
>> Another possibility is that the ATA controller to which the drive is
>> connected is faulty. This one is easy enough to rule out: swap both
>> hard drives and see if the problem persists (the "disk0s3" portion of
>> the error message in system.log will change, I believe - but the error
>> will be the same otherwise).
>>
>> Ruling those two out, another possibility is that some part of the
>> drive hardware is faulty - the drive motor, read/write heads, circuit
>> board, or firmware is faulty. If you've gotten this far, I recommend
>> you run Disk Utility to see the SMART status of the drive. If the SMRAT
>> status is verified, then run TechTool Pro (or Deluxe, if you have
>> AppleCare) and see if it can tell you any more about the drive.
>> Ultimately, if you've gotten this far, you're looking at replacing the
>> drive. Is the drive under warranty?
>
> Drive is over a year old.... but the SMART readings from Disk Utility
> say 'verified' for both the drives. TechTool Pro is running in the
> background now, and is half way through. I had already run it on
> everything but the surface scan.
>
> Since I have been thinking about a new drive anyway, putting in a new
> one might be simplest, and thus backing up and reformatting the
> potentially dodgey drive would be easiest. Swapping the drives might be
> doable on the G5, but do I have to change any settings on the drives?
> Or does the one in the "A" Bay (where the original drive sits)
> automatically become Master? Or was that done away with with Serial ATA
> drives. I can't recall if I set any pins on that second drive I put in
> rather hurriedly.

Nope there are no settings required for SATA. Just swap them and boot
back up and see if you get the problem on the other drive or on the
same drive.


--
JR

On 2006-12-20 15:31:56 -0600, russotto@grace.speakeasy.net (Matthew T.
Russotto) said:

> In article <W5GdnakA7pryNBTYnZ2dnUVZ_qisnZ2d@comcast.com>,
> John McWilliams <jpmcw@comcast.net> wrote:
>>
>> Drive is over a year old.... but the SMART readings from Disk Utility
>> say 'verified' for both the drives. TechTool Pro is running in the
>> background now, and is half way through. I had already run it on
>> everything but the surface scan.
>
> If you have a more detailed SMART utility, check the UDMA CRC error
> count. I had a drive (brand new, a Western Digital Caviar) which
> didn't fail SMART but kept getting DMA CRC errors; I assume either a
> bad or incompatible controller. Symptoms were the same -- timeouts
> and failures.

SMARTmontools will give you *very* detailed information. I believe
I've installed and run this in Mac OS X in the past - it was a simple
compile from the source, IIRC. You can get the source here:

<http://smartmontools.sourceforge.net/>

Installation instructions are on the same web page under the heading
"Second Method (Linux/Solaris/FreeBSD/NetBSD/OpenBSD/Cygwin) - Install
from the source tarball". Basically, you will:

1. Download it from here:
<http://sourceforge.net/project/showfiles.php?group_id=64297>
2. Expand the tar ball (double-clicking it should expand it).
3. Open Terminal.
4. Enter: cd (the path to expanded folder)
5. Enter: ./configure
6. Enter: make
7. Enter: make install

--
JR

On 2006-12-20 18:47:09 -0600, Jolly Roger <jollyroger@null.org> said:

> 7. Enter: make install

Make that:

7. sudo make install

: )

--
JR