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

Given the fact that most probably my previous issue is due to a hardware failure: Linux Arch on ARM - Random USB drive "freeze"

It seems that the malfunctioning started when I had a lot of torrent seeding with Transmission (which has no queue support); I usually (for ethical reasons) keep seeding until seed ratio is 1.1 and have an internet connection with download/upload ratio of about 20, so usually time of seed is several times the time of download. This leads to the drive being active almost 24/7.

Is this really dangerous for my hard drive? Or did I just have bad luck?

It is ok to me to replace the bittorent hard disk every 2-3 years, where the one I dismiss goes to samba sharing or completely dismissed.

Edit: a few details to help understand the scenario.

The computer is actualy a plug computer with only 128MB of RAM (not expandable); the files usually are 300-700 MB large so no cache can help here.

During normal work it seeks sometimes, say every few seconds at most, not continually; my use is not worth a server class HD.

SMART says the faulty drive is perfectly fine!

Reported Google study shows the fact that my hd is probably somewhere in the area of the ones which fail in the first year of life... bad luck!

I actually use Deluge now because it has queue support.

The drive is not dead, it seems that it "takes a pause" unpredictably, seems more often as the average load (mostly repeated seeks) raises.

share|improve this question
up vote 3 down vote accepted

You can read an interesting study done by Google into HDD failure here.

As far as age goes, Google's data show drives start to fail more often at the 2 year mark, but adding the columns you can see that after 5 years there is still only a 35-40% total chance of failure.

I dont think your seeding will cause increased wear on the HDD, but it probably depends on how many files are being seeded and to how many peers - that would affect how often the drive heads have to move to read the next chunk of data. A certain amount of data is usually cached in memory on the computer (which reduces the amount of work the HDD must do), I'm not sure if that is configurable in Transmission, or how much free memory your device has to use for caching.

Try listening to the drive with a typical workload running (number of torrents/peers). If you hear constant mechanical sounds (other than the spinning), that means its reading/seeking all the time, which is arguably worse than idle spinning. If it only makes noises occasionally, then there's not really anything to worry about.

One option is to keep using the same drive with the seedbox until it dies, and use new drives for archival/backup purposes (move the files onto them once enough seeding has been done). Then if it dies you only lose current torrent downloads/seeds. Of course if the new drives are in a computer or NAS, they could be RAIDed for some level of redundancy if the data is really important to you.

share|improve this answer
this is the most informative response, thanks; I guess I can keep going like this and hope I don't get another faulty drive! – Riccardo Cossu May 20 '11 at 12:41

Hard disks should have an estimate life of 60000 hours, therefore replacing it every 2-3 years should be safe.

In addition, predictable failures can be prevented with SMART tools:

share|improve this answer
it's almost 7 years :-) so I guess even 5 should mostly be ok. But my question is a bit more specific, does bittorrent (in particular seeding) put the hard disk in such stress 24/7 that may lead to failure more often than predicted? – Riccardo Cossu May 19 '11 at 11:41

There's no way to answer your question. How much RAM cache does your disk have? How big is the single file you seed at any one time? See, if the whole file fits in RAM and you aren't using the disk for anything else, the actual drive hardware won't be hit at all by any uploading of that file. OTOH, seeding dozens of large files would certainly mean exercising the mechanism quite a bit.

Modern hard drives are quite robust and I wouldn't expect problems. If you were going to be continuously uploading/seeding 50 files every day, you might consider investing in a server-class drive which would be designed for continuous use with long MTBF.

share|improve this answer
thanks for your feedback, a server-class drive is an overkill for my use :-) – Riccardo Cossu May 20 '11 at 12:40

You must log in to answer this question.

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