Unexpected news

  • Thread starter Thread starter Joe S
  • Start date Start date
J

Joe S

Here are a few in-context snippets from a recent report

"Hard disk test 'surprises' Google"
Monday, 19 February 2007, 15:28 GMT
http://news.bbc.co.uk/1/hi/technology/6376021.stm




The report said that there was a clear trend showing "that lower
temperatures are associated with higher failure rates".

"Only at very high temperatures is there a slight reversal of this
trend."

But hard drives which are three years old and older were more likely to
suffer a failure when used in warmer environments.

"This is a surprising result, which could indicate that data centre or
server designers have more freedom than previously thought when setting
operating temperatures for equipment containing disk drives," said the
authors.
 
Joe said:
Here are a few in-context snippets from a recent report

"Hard disk test 'surprises' Google"
Monday, 19 February 2007, 15:28 GMT
http://news.bbc.co.uk/1/hi/technology/6376021.stm

The report said that there was a clear trend showing "that lower
temperatures are associated with higher failure rates".

"Only at very high temperatures is there a slight reversal of this
trend."

But hard drives which are three years old and older were more likely to
suffer a failure when used in warmer environments.

"This is a surprising result, which could indicate that data centre or
server designers have more freedom than previously thought when setting
operating temperatures for equipment containing disk drives," said the
authors.
Click to expand...

I don't believe a word of it.

The majority of drives I receive for recovery have indications of
burnt-out chips or other markers pointing to heat damage.

Heat can also corrupt a drive's firmware.

I receive far more drives that have been installed in external housings
(with no active cooling) than otherwise.

Not only that, but "data centre" or "server designers" tend to use SCSI
drives, which are infinitely more robust than the average PATA / SATA
drive.

Sorry, but their findings are flawed.


Odie
 
I don't believe a word of it.

The majority of drives I receive for recovery have indications of
burnt-out chips or other markers pointing to heat damage.

Heat can also corrupt a drive's firmware.

I receive far more drives that have been installed in external housings
(with no active cooling) than otherwise.

Not only that, but "data centre" or "server designers" tend to use SCSI
drives, which are infinitely more robust than the average PATA / SATA
drive.

Sorry, but their findings are flawed.
Click to expand...


Their findings are what they are.

Unless you can point to an arithmetic error or an error the
methodolgy, how can you use the word, "flawed'.
 
Al said:
Their findings are what they are.

Unless you can point to an arithmetic error or an error the
methodolgy, how can you use the word, "flawed'.
Click to expand...

They contrast significantly with the findings I have - mainly the "heat"
issue.

Granted, I don't have the turnover of drives that they do - but my
experience tells me there is something rather suspect (although probably
not at all intentionally) in their findings.


Odie
 
Those are in fact temperatures not normally seen in personal desktop
systems, like 20C. And while Table 4 shows a quite pronounced effect,
Table 5 doesnt, and only much effect with very new drive failures anyway.

Table 5 doesnt really show that either, its showing drive temps, not environmental temps.

The report actually found that there isnt any significant difference between
SCSI, FC and SATA drives reliability wise apart from one bad SATA design.
Their findings are what they are.
Click to expand...

Not necessarily. Since their's is the only 'study' to come up with that result,
and that is in contrast to what all the rest have found, the other obvious
possibility is that it isnt a real effect and its just the result of some other
variable like the drive models used etc and doesnt even say anything about
what happens when the SMART temp is routinely above 50C etc either.
Unless you can point to an arithmetic error or an error
the methodolgy, how can you use the word, "flawed'.
Click to expand...

Easy when their claim doesnt match what others have found.

Its MUCH more likely to be an artifact instead, particularly
when that conclusion doesnt gell with the basic physics either.
 
Odie Ferrous said:
Granted, I don't have the turnover of drives that they do - but my
experience tells me there is something rather suspect (although probably
not at all intentionally) in their findings.
Click to expand...

Their data sample only covered nine months. I don't think that is long
enough to draw any conclusions about whether heat affects drive
reliability or longevity.

If the survey had covered a period of three years, and had ignored data
imported from the previous maintenance database, I suspect the results
would be very different.
 
If a chip burns out, it's more likely caused by tiristor effect caused by
surge on the output pins or ESD. It can be caused by extreme overheat, too,
but it's not what Google paper talks about.
 
Previously Alexander Grigoriev said:
If a chip burns out, it's more likely caused by tiristor effect caused by
surge on the output pins or ESD. It can be caused by extreme overheat, too,
but it's not what Google paper talks about.
Click to expand...

Actually a visually detectable burned out chip is very unlikely
to be caused by overheating. It would need to heat up to something
like 250C for that. A visibly burned out chip is usually
voltage surges or badly manufactured chips that short themselves
out. Not external heat.

Arno
 
Arno Wagner said:
Actually a visually detectable burned out chip is very unlikely
to be caused by overheating. It would need to heat up to something
like 250C for that. A visibly burned out chip is usually
voltage surges or badly manufactured chips that short themselves
out. Not external heat.
Click to expand...

There is some evidence that particularly with Maxtors, running the drive
too hot can get it to fail and that failure does result in a visibly fried chip.

 
Arno Wagner said:
Actually a visually detectable burned out chip is very unlikely
to be caused by overheating.
Click to expand...
It would need to heat up to something like 250C for that.
Click to expand...

And that obviously doesn't qualify as 'extreme', eh, babblebot?
A visibly burned out chip is usually voltage surges
Click to expand...
or badly manufactured chips that short themselves out.
Click to expand...

Because of overheating, e.g.
Not external heat.
Click to expand...

Helped by external heat.
Arno
Click to expand...

[utter crap snipped]
 
Back
Top