OT: Is any hard disk brand better/worse than the others?

[Lennart Sorensen]

On Mon, Dec 01, 2008 at 12:17:49PM -0500, D. Hugh Redelmeier wrote:
> Lots of people have opinions.  Only a few have a large enough sample
> size to do a rigorous study (eg. Google) and they aren't talking.
> 
> The people willing to express definite opinions are probably the ones
> who jump to conclusions.
> 
> Anecdote:
> 
> IBM was a very good builder of hard drives.  They blew one generation
> ("Deathstar") and they didn't know until a *lot* were in the field.
> Their reputation was hurt so badly that they sold the business to
> Hitachi (remember: IBM invented the disk drive!).  Even in Hitachi's
> hand's, the reputation sticks.  (BTW, the second last drive I bought
> was a Hitachi to upgrade my ThinkPad; one reason I chose Hitachi was
> that the drive was replacing a Hitachi.)
> 
> Henry gave me a couple of the bad Deskstar drives, unused.  Apparently
> there is no recourse at this late date.
> 
> I think that each generation from each manufacturer is a crap shoot.
> By the time you have a feeling about a drive, the news is obsolete.
> 
> Having said that, I am leaning towards the Spinpoint 1T drives at the
> moment.  My basis is chit-chat that I'd not put any stock in if I had
> real data.
> 
> The most recent hard drive that I bought was a WD 1T drive in an
> external package, sold by Dell at a very good sale price.  Price was
> one reason.  I needed eSata, and cheap boxes often don't do that.
> This was a GP disk, using less power (5400RPM, a detail they try to
> hide) -- the right tradeoff for my particular application (expand a
> Rogers HD PVR).

Well the 5400rpm drive actually has no problem keeping up with USB so
for most external enclosures it is fine.  For eSATA though, you would
notice the difference.  They probably expect most people to just use USB
and hence not care.

> I've heard bad things about the 1.5T Seagate drives but I don't trust
> those reports.  Apparently there is yet another limit on disk sizes: 2T for
> SATA, but a common Windows driver bug (signed vs unsigned?) drives
> that limit down to 1T.  I've not done the research to know if those
> limits are real or come from people who just make stuff up.

Ehm, DOS style partition table has a limit of 2TB (2^32 sectors).  Might
be able to stretch that to 4TB if you created a 2TB partition ending
just before the 2^32 sectors, and another 2^32sector partition starting
just before, although I am not sure that is a legal partition table.
The start offset is 32bit and the size is 32bit, so who knows.

Using GPT (EFI style partitions) solves that problem.  I currently use
that for some 2.3TB array volumes.

Of course windows isn't likely to support that yet, so you could have a
problem there.

-- 
Len Sorensen
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists