[mythtv-users] Re: Hard Drives that Actually Work?

[Chris Petersen]

> 1 / I have been told by two large storage vendors that they "cherry 
> pick" the best SATA drives for their Enterprise Disk Arrays.  You can 
> read this in various ways, I take it to mean that while hard-drives past 
> testing to be sold, they are not all equal (similar to testing and speed 
> rating CPU's)

Any enterprise-level vendor worth their weight will perform some serious 
burn-in tests on systems they sell.  We use cerberus 
(http://freshmeat.net/projects/ctcs/) and burn for no less than 24 hours 
  (usually 2-3 days).  It's extremely rare than anything will pass 
through burn-in and then fail in the field under normal use.

The other thing to consider is that with drives, they usually go bad in 
lots.  So when we get a bad drive, we usually go through and do extra 
testing on all of the other drives from that lot (since about half the 
time, there's something bad in the whole lot).

-Chris