[mythtv-users] How is this pricing possible?

[John Drescher]

On Wed, Feb 23, 2011 at 5:50 PM, Alex Butcher <mythlist at assursys.co.uk> wrote:
> On Wed, 23 Feb 2011, Rob Smith wrote:
>
>> On Wed, Feb 23, 2011 at 2:26 PM, Alex Butcher <mythlist at assursys.co.uk> wrote:
>>> That's probably because some of those raw values are actually multiplexed
>>> values, and you need to use a binary mask to demultiplex the different
>>> original attributes:
>>>
>>> <http://forums.seagate.com/t5/Barracuda-XT-Barracuda-and/Interpret-Raw-SMART-Data-Read-Error-Rate-amp-HardWare-ECC/m-p/45617#M18267>
>>>
>>> Best Regards,
>>> Alex
>>
>> I just don't trust a random forum user over Seagate's official
>> knowledge base. Sorry.
>
> Well, as I wrote previously, you don't really need to - just compare the
> cooked value against the threshold.  If it's less than the threshold, the
> attribute is considered to be in a 'failed' state, and the official Seagate
> tool should confirm this - if necessary - for warranty purposes.
>
> The raw values are only really useful for Seagate and out of curiosity for
> the likes of us ("how failed is my drive?  a) barely b) somewhat c)
> totally screwed!")
>

Hmmm. For all 4 manufacturers, I only look at raw values and pretty
much ignore the thresholds and cooked values.. At work I have nagios
monitoring dozens of drives for several different raw values so I get
alerts on my desktop a few minutes after one of these changes. When I
see things like reallocated sectors growing daily/weekly I kick the
drive out of whatever raid5/6 array it is in and then test it with a 4
pass badblocks r/w test. If the numbers change drastically after this
I repeat and most of the time after that the drive will be bad enough
to fail the manufacturer tools. Then I RMA the drive.

John