Curious temperature / transfer speed behaviour

More and more it seems the HD in the iMac slows down in ability to read as HD temperature goes up.
Alan Browne wrote on :

In the continuing saga over the odd hanging behavior of my iMac...

Three temperature points done over two cycles (last night and today).

CPU temp 30 - 40C during tests with a CPU load of 2 - 4% on each core over the course of the transfer; occasional jumps to 10 - 20%.

Test: transfer a large (closed) file from the system disk to a backup disk via Firewire. Same file is transferred each time.

HD temperature Time to transfer one 43 GB file (mm:ss)

1st round 28 - 38C (rose 10Ú during test) 13:50

48C 22:51

50C 24:11

2nd round (machine cooled overnight (Sleep)). 30 - 39C (rose 9Ú during test) 14:01

48C 23:09

50C 24:08

More and more it seems the HD in the iMac slows down in ability to read as HD temperature goes up.

Could be the internal read error rate is higher?

The disk speed is lower as temperature rises?

Smart status is "verified".

Jolly Roger replied on :

In article itudnaEHxJMfRCnRnZ2dnUVZ_j6dnZ2d@redacted.invalid, Alan Browne alan.browne@redacted.invalid wrote:

In the continuing saga over the odd hanging behavior of my iMac...

Three temperature points done over two cycles (last night and today).

CPU temp 30 - 40C during tests with a CPU load of 2 - 4% on each core over the course of the transfer; occasional jumps to 10 - 20%.

Test: transfer a large (closed) file from the system disk to a backup disk via Firewire. Same file is transferred each time.

HD temperature Time to transfer one 43 GB file (mm:ss)

1st round 28 - 38C (rose 10Ú during test) 13:50

48C 22:51

50C 24:11

2nd round (machine cooled overnight (Sleep)). 30 - 39C (rose 9Ú during test) 14:01

48C 23:09

50C 24:08

More and more it seems the HD in the iMac slows down in ability to read as HD temperature goes up.

Could be the internal read error rate is higher?

The disk speed is lower as temperature rises?

Smart status is "verified".

It could be one o the semiconductor components on the drive is experiencing a failure caused by overheating.

What model hard drive is this? What is the manufacturer's stated operating temperature range for that model?

Alan Browne replied on :

On 10-10-13 9:31 , Jolly Roger wrote:

It could be one o the semiconductor components on the drive is experiencing a failure caused by overheating.

What model hard drive is this? What is the manufacturer's stated operating temperature range for that model?

WDC WD5000AAKS-40YGA0 (500 GB Western Digital).

With another cycle of tests I've noticed the "freeze" correlates more closely to the temperature of the HD bay than the HD temperature itself. (Unless the HD is reporting temperatures late because of the issue itself?)

So I'm wondering if it's more of a mechanical issue (thermal expansion affecting a connection or something of that nature) than the HD itself.

Now I got the "freezes" to occur again whenever the HD temp crosses 51C.

I DL'd a program called SMART Utility. It reports failures whenever the HD is in this hanging mode. (Apple Disk Utility reports the disk as "verified" even when the computer is in the intermittent hang state ...)

The spec for the drive says operating temperature of 0 to 60C.

Per Google, page 6 of this report, http://labs.google.com/papers/disk_failures.pdf it appears that the best temperature range to operate a HD is about 35-45C. (If I'm interpreting that right).

Given the "failing" report from SMART Utility and that reports only occur when in the "hang" zone, I'll run the disk in the range above by manually controlling the fan (if I keep it at 1200 rpm that seems enough for most of my use).

As the computer is still under Apple Care, will they say "there's no problem, the SMART status is 'verified'" or will they accept my report and the report from SMART Utility?

Will they accept putting in a disk that I supply to replace the failing one?

nospam replied on :

In article w_OdnZlIQr-LkyvRnZ2dnUVZ_gCdnZ2d@redacted.invalid, Alan Browne alan.browne@redacted.invalid wrote:

As the computer is still under Apple Care, will they say "there's no problem, the SMART status is 'verified'" or will they accept my report and the report from SMART Utility?

they might. i'm sure they know that disk utility's smart check is minimal. they may even have their own tests to verify it.

Will they accept putting in a disk that I supply to replace the failing one?

not under warranty they won't.

Jolly Roger replied on :

In article w_OdnZlIQr-LkyvRnZ2dnUVZ_gCdnZ2d@redacted.invalid, Alan Browne alan.browne@redacted.invalid wrote:

On 10-10-13 9:31 , Jolly Roger wrote:

It could be one o the semiconductor components on the drive is experiencing a failure caused by overheating.

What model hard drive is this? What is the manufacturer's stated operating temperature range for that model?

WDC WD5000AAKS-40YGA0 (500 GB Western Digital).

With another cycle of tests I've noticed the "freeze" correlates more closely to the temperature of the HD bay than the HD temperature itself. (Unless the HD is reporting temperatures late because of the issue itself?)

So I'm wondering if it's more of a mechanical issue (thermal expansion affecting a connection or something of that nature) than the HD itself.

Now I got the "freezes" to occur again whenever the HD temp crosses 51C.

I DL'd a program called SMART Utility. It reports failures whenever the HD is in this hanging mode. (Apple Disk Utility reports the disk as "verified" even when the computer is in the intermittent hang state ...)

The spec for the drive says operating temperature of 0 to 60C.

Per Google, page 6 of this report, http://labs.google.com/papers/disk_failures.pdf it appears that the best temperature range to operate a HD is about 35-45C. (If I'm interpreting that right).

Given the "failing" report from SMART Utility and that reports only occur when in the "hang" zone, I'll run the disk in the range above by manually controlling the fan (if I keep it at 1200 rpm that seems enough for most of my use).

As the computer is still under Apple Care, will they say "there's no problem, the SMART status is 'verified'" or will they accept my report and the report from SMART Utility?

Will they accept putting in a disk that I supply to replace the failing one?

Best to ask them.

They will want to reproduce the problem themselves, which will require you to tell them how to reproduce it and give them your machine so they can.