I was curious if people here pre-emptively replace hard drives based on usage or time in service, or do you wait for signs of problems to show up before replacing them? 

 

If done pre-emptively, what criteria do you use?

Interesting question..

 

With SMART these days. You usually get a heads up before disaster...

 

Unfortunately.. I have and still do loose drives with no warning. I tend to keep duplicaes of most of my data and run them untill i get smart status errors or full on death. what ever comes first.

 

I did have a drive recently that was un reliable. i got CRC errors and lost data on it and it passed smart and i RMA'd it.

 

The drives that don't fail in 3-5 years tend to end up in my junk pile. they are to small and to slow for my needs (not to mention probably energy inefficient). the drive i had 3-5 years ago were probably 500 gig (ide and sata mix) or the early LP 1TB drives.

 

I tossed all IDE drives at this point

 

The older 500gb sata2 drives i use for boot drives in windows servers.

the 1TB drive i gave some away. put back ups of stuff on them and shelved some of them (they are prolly not going  to work if i need them). I also put some of them in USB/ESATA enclosures for moving data from place to place.

 

The 1.5TBs i have will go into my second unraid box. possibly some older 1tb drives and ear mark them for replacing...

 

I'll guess in 3-5 years we will see  bigger  and or faster hard drives. possibly massive, cheap SSD's. again making my current 2TB and 3TB drives obsolete before failure...

 

then again.. it seems like hard drive quality is going to crap.. i have had very few 500gig or smaller drives die while i have had lots of 1 TB and up die.

I was curious if people here pre-emptively replace hard drives based on usage or time in service, or do you wait for signs of problems to show up before replacing them? 

 

If done pre-emptively, what criteria do you use?

 

This is why I normally buy 5 year warranty drives. I know this may not be the most moral thing to do, but before the 5 year warranty is up I'll do a warranty process and get another drive. ALTHOUGH by that time the drive sizes will be small, newer and better technology and different interfaces. I've only done this several times. Half the time it is not worth it because of the drive sizes are so small.

I was curious if people here pre-emptively replace hard drives based on usage or time in service, or do you wait for signs of problems to show up before replacing them? 

 

If done pre-emptively, what criteria do you use?

 

This is why I normally buy 5 year warranty drives. I know this may not be the most moral thing to do, but before the 5 year warranty is up I'll do a warranty process and get another drive. ALTHOUGH by that time the drive sizes will be small, newer and better technology and different interfaces. I've only done this several times. Half the time it is not worth it because of the drive sizes are so small.

 

Caveat venditor!

 

I don't do this. What I have found, though, is at around the 3-5 year mark, larger drives are available to replace the old ones at a very low pricepoint. For example, I replaced 3 750G drives with a 2T drive about a year ago for $60. No brainer IMO.

 

But if this golden age of disk drive size increases/price decreases is nearing it's end, I'll be monitoring my older disks' smart reports for tell-tale signs of failure and replacing them as indicated. 

I was curious if people here pre-emptively replace hard drives based on usage or time in service, or do you wait for signs of problems to show up before replacing them? 

 

If done pre-emptively, what criteria do you use?

 

This is why I normally buy 5 year warranty drives. I know this may not be the most moral thing to do, but before the 5 year warranty is up I'll do a warranty process and get another drive. ALTHOUGH by that time the drive sizes will be small, newer and better technology and different interfaces. I've only done this several times. Half the time it is not worth it because of the drive sizes are so small.

 

I used to do this in the days of Maxtor.

 

I find I have not done this in quite some time since drive sizes double or triple by the time I take a drive out of service.  For example. I still have 16 WD 1tb drives in service. None have failed yet. 

In comparison 2 WD 2TB drives have failed and 1 Seagate 1TB drive has failed.

 

After a recent cleanup

I found I still had

5 250-320GB SATA drives in spare storage.

7 500GB SATA drives in spare storage.

3 750GB SATA drives in spare storage.

5 320GB IDE drives in spare/refurbished storage.

4  80GB IDE drives in spare /refurbished storage.

 

Too bad I now need some 500GB 3.5" pata and 320GB 2.5" pata drives for older machines.

 

I'll guess in 3-5 years we will see  bigger  and or faster hard drives. possibly massive, cheap SSD's. again making my current 2TB and 3TB drives obsolete before failure...

 

then again.. it seems like hard drive quality is going to crap.. i have had very few 500gig or smaller drives die while i have had lots of 1 TB and up die.

 

This is a good point.  I think about 3 years ago I was buying 250 -500 GB drives for what I'm now buying 2 TB drives for.  I'm not sure it will increase at the same rate, but in 2-3 years I'm sure the playing field of drive availability will be different.  

 

I don't deal with the number of drives going through my hands that a lot of you do, but I have bought 3 Hitachi - 2 TB hard drives and one of them had to be RMA'd twice as DOA.  That'd 3 out of 5 bad, 40%.  Not exactly a good pool to work from statistically speaking, but to have the replacement come in DOA as well is not good.  It seems the user reviews for just about any drive today seem to have a large number of people talking about DOA issues.

 

I wonder if it has more to do with build quality, or the end results of trying to squeeze so much data on such a small drive that still has mechanical components.  

 

I've been through 3 or 4 other hard drive failures either at home or work.  It only took the first one to understand the importance of a backup plan (and this was a number of years ago).  They were all older drives though that could easily be explained due to age /wear, not still in the brand new hardware stage.