sitotisak said:
just picking the right subject,
i'm trying to find out a little more about wearability of sata disks
on different acess and output speeds.
Do sata disks which are fully using their speed capability ie. at
their optimal speed of 3gb/s(in this case) do wear more and sooner
than the same capacity disk off lesser speed i.e. 1.5Gb/s ?
In what atributes in SMART tool could that be recognised?
thanks for writing!
First, you would start by researching all the components in
the hard drive that wear. Then, check to see which of those
is affected by the cable rate.
My guess would be, the cable rate makes almost no difference
(using current disk technology).
The drive does use "flex cable", something like a Kapton film
with conductors on it. That cable flexes each time the head
moves. Some people have posted in the USENET groups, that
their server hard drives last about a year, if pounded by
many small operations per second. But I haven't seen any claims
as to what is killing those drives. You'd need to find
some academic papers that analyze drive failures, to see
what the issue might be.
(This paper is no longer available, but is a bit interesting.)
http://www.siliconsystems.com/technology/pdfs/Hard_Drive_Failure_Studies.pdf
"Moreover, the HPC3 data set includes both SCSI and SATA drives
(as part of the same system in the same operating environment) and
they have nearly identical replacement rates. Of course, these
HPC3 SATA drives were decommissioned because of media error rates
attributed to lubricant breakdown (recall Section 2.1 on page 3),
our only evidence of a bad batch, so perhaps more data is needed
to better understand the impact of batches in overall quality."
The SATA and SCSI drives wouldn't have identical IOP rates, and yet in
that experiment, replacement rates were about the same.
The contents of that paper, are also available here.
http://www.cs.cmu.edu/~bianca/fast/index.html
Google did a study on disks, but it isn't interested in quite
the same issues as that paper.
Paul
