• Register

What is the service life of an external hard drive?

+3 votes
752 views
Many institutions and groups use external hard drives for backups. Are there any good practice guidelines for these devices? For instance, is it always better to unplug the hard drive when not in use? And if the rule-of-thumb is 3-4 years between replacements on internal disks, is there one for replacing external drives?

Edit: I'm asking on behalf of another listserv and as part of my own search for equipment replacement advice.

I've seen many digitization projects where external hard drives that were purchased for use in a project years ago are still in service as because they haven't failed yet. For instance, I helped with a digitization project using a Firewire 800 external drive 3 years after computer manufacturers stopped supporting that interface. Often these hard drives lie fallow for months between projects, and then dusted off to be pressed into service on request. Even if the data is backed up from these drives, they are often the working storage for a project until the scheduled backup. At what point should a librarian/archivist/etc stop trusting that a drive will last the duration of a project, even if it has all the right specs for the job?
asked Jun 5, 2014 by nkrabben (1,990 points)
edited Jul 1, 2014 by nkrabben

3 Answers

+3 votes
 
Best answer

This question has been answered here

The basic conclusion is that, apart from keeping the drives in a recommended temperature range, I'm not sure that there is any great way to prolong the use of the hard drives. There has been reseach into how often hard drives last when not spun-up regularly vs when spun-up regularly but it appears mostly inconclusive. 

Regardless, it might be better to ask what is the best way to prolong the use of the content on the hard drives. And the best answer in this context (the context of bit preservation) is to make more copies. If you have done that then if a hard drive holding one copy fails then you will still have access to the other copies of the content. 

answered Jun 27, 2014 by euanc (3,910 points)
selected Jul 15, 2014 by nkrabben
Thanks for the link. I updated my question with a bit more specific use case, namely when to take external drives out of service that have been passed from project to project for years.
I would extend “make more copies” to “make more copies and regularly verify them” — hard drive failures aren't evenly distributed and there are plenty of other sources of error but you can often see significant service lifetimes as long as you have a good verification workflow.
+3 votes
If it is really a "drive" with moving parts, don't move it around! We see many many hard drives in student laptops biting the dust all the time because they are banged around, while a hard drive in a laptop that sits on a desk for its whole life doesn't have this problem. Same (especially) goes for external HDs, which look so sturdy but if not solid state are potentially fragile.
answered Jun 27, 2014 by patgalloway (200 points)
+1 vote

Just came across this thread as someone in my organisation came to me with a similar question. As for guidelines for the use of hard disks, the WikiBook below contains some useful additional suggestions:

http://en.wikibooks.org/wiki/Minimizing_Hard_Disk_Drive_Failure_and_Data_Loss/Environmental_Control

One thing covered here that isn't mentioned in the other answers, is the presence of strong magnetic fields. This is definitely something to check for if your disks are stored in the vicinity of machinery.

answered Aug 25, 2014 by johanvanderknijff (2,060 points)
Does this include being near a power strip, especially one with multiple transformers? I would imagine it does, but I couldn't find any good sources.
Couldn't find much on this either. I don't know if this helps, but I did come across a few sources on DIY methods for approximate measurements of magnetic field strength. Might be worth a look / try to get a rough idea:

http://physics.stackexchange.com/questions/46292/rough-easy-diy-method-of-measuring-magnetic-field-strength

And:

http://www.wired.com/2014/01/measure-magnetic-field/

(BTW I've never tried any of these methods myself, and I have absolutely no idea how useful/reliable they are for this purpose!)
...