Pre-emptive drive replacement?


tj80

Recommended Posts

Hi,

 

My UnRAID box (6.4.0 currently) has been running nicely for many years now, including saving me from one failed WD Green 2TB data drive.  However, some of the drives (the parity disk in particular) are getting pretty old and I wonder what the consensus is on whether to pre-emptively replace drives or just wait until they fail?

None of my drives have any SMART warnings or other issues I know of (no funny noises, high temperatures, etc), they auto spin down so haven't been running continuously, and see very light use in a domestic environment - but in terms of power on hours some are getting pretty old.  Yes I have backups, but still don't want to lose anything!  I'm not so concerned about losing one drive obviously, more that the intensive workload recalculating parity after a single failure might kill off a second drive at the worst possible time.

 

Power-On Time:

Parity - 2TB Hitachi Deskstar, 7y 9m

Data 1 - 2TB WD Green, 6y 7m

Data 2 - 2TB Seagate Barracuda, 3y 7m

Data 3 - 250GB HP Enterprise SATA, 5y 8m

Data 4 - 500GB Seagate Momentus (2.5"), 1y 8m

Cache - 256GB Crucial M4 SSD, 2y 6m

 

Most are well past the "infant mortality" timeframe so are clearly good solid drives - so should I just keep them running until they fail or think about replacing early?  One option I thought about was to replace the old 2TB parity disk with a new one, then use the old drive to replace the 250GB Data 3 - but I don't actually need any more space yet!

 

What would you do?

 

Many thanks,
Tim

Link to comment

Personally, I would run them until you start to see some issues.  Just be sure that you have notifications setup so you can see the first signs of a problem.  And run monthly parity checks.  Think carefully about developing a plan for upgrade/replacement as you need more space and a failure actually occurs.  

Link to comment

I agree with @Frank1940.  If you are running monthly parity checks, then your sytem is already known to be capable of reading from all of the drives at once.  The only difference when recovering from a failed drive is that one of those drives being read becomes a drive being written to.  The loading on the sytem is the same. 

 

If I was to replace any part pre-emptively I would go for the power supply.  They do reduce their ability to provide maximum power as the capacitors within get older.  A decent power supply run conservatively would most likely give no trouble when in use continually for, say, eight to ten years, but a failing power supply will potentially give rise to strange problems some of which may be hard to detect and which might be more challenging to recover from.

Link to comment

Thanks, both - reassuring!  Think I'll leave it be (yes I have automatic email alerts for failures set up).

 

Interesting point about the PSU, to be honest I hadn't thought about that.  It's an HP Microserver which has been running 24x7 for 7 years, so the PSU should be decent quality and it's not too heavily bogged down (4 x 3.5" disks, 1 x 2.5" disk and an SSD - no expansion cards, etc).  Will definitely keep it in mind though.

 

Cheers,
Tim

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.