Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Running a RAID 0 setup and was wondering if running defrag too much would lead to a disk failure faster? Disks fail based on doing too many reads and writes? or is it based on the spinup and spindowns?

share|improve this question
Curious as to why you see the need for daily defrag anyway. Unless you are moving huge amounts of data onto/of of the disk, fragmentation isn't likely an issue. – uSlackr Apr 20 '12 at 15:16
up vote 7 down vote accepted

Google's study on disk failures is to my knowledge the largest available research from real-life data concerning disk failures. They found a surprisingly small correlation between utilization/temperature/power cycle counts and disk failure. That is, there are other variables that have a larger influence.

Google found that when some specific SMART values started to increase, the drive had a 40 times greater risk of failing within two months as compared to a "healthy" drive, but when working on such a minimal set as one does in a home environment, this is not really anything one can draw any real conclusions from. The sample size is simply too small to draw raw conclusions on almost all parameters for home PC:s concerning disk failure (well, don't let them endure physical stress (drop them, etc.)).

"Is it based on", you say. Well, perhaps to some extent, but probably so small that it is not noticeable in such a small sample.

share|improve this answer
That's a nice study of Server drives... but it doesn't really help those home users... a you pointed out. In fact, when you said "when working on such a minimal set as one does in a home environment, this is not really anything one can draw any real conclusions from" you kind of said that your whole answer was a big "I don't know". – Bon Gart Apr 20 '12 at 15:12
That is a nice study. Will be sure to read it. – Tiffany Walker Apr 20 '12 at 15:17
@BonGart: "I don't know" would have been to say "I don't know what impact the parameters have". I say "The parameters have such a small impact that you can forget about them in such a small sample size", which seems to be the scientific consensus. The problem is when forums are filled with personal hunches and things people have "heard", and "it stands to reason that..." and so on, combined with a lack of understanding of statistics. For some reason this is extremely common among hardware sites (for *deity*'s sake, don't mention laptops and batteries... :-D ). – Daniel Andersson Apr 20 '12 at 15:27
"said that your whole answer was a big "I don't know" My interpretation is Daniel said nobody knows for sure. – Moab Apr 20 '12 at 15:50

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .