While I am confident that I am not the only person this has happened to, I wanted to share my story for anyone who still thinks it is a good idea to keep spare drives sitting around for unraid.  I probably built my system 6 years ago (hard to say exactly).  I was on unraid 4.7 and we were getting close to the big Thailand flood that would cause HDD prices to skyrocket.  Whenever I caught drives on sale, I would buy them.  I figured they would be good to use to with expand storage or replace a failed drive.   I personally had two 2TB drives sitting unopened in my rack.  

 

The first drive was actually an RMA from WD for a failed drive some time ago.  One of my older 2TB drives failed last week, so I obviously had the perfect solution.  Cracked the plastic, precleared the drive and rebuilt from parity.  I love it when a plan comes together....    7 days later,(yesterday) this drive fails out of nowhere.  I figured it must be a cabling issue, so I re-seat the drive and rebuild again...fail...   Of course, I call WD and explain the situation and I was unsurprisingly told that they would not replace the drive because it was out of warranty.  They would not accept smart reports showing that I had only used the drive for a few days.

 

Not the end of the world right?  I still have a brand new 2TB Drive in the box.  I pick up the box and the first thing I notice is the price tag.  $99.99....   I crack the plastic, preclear the drive and rebuild.  So far its at 80% with no issues.  However, it really made me think.  Today, for that same $100, I could buy a drive in the 4-5TB range.  Lets not forget that the 2TB that I just opened faces the same previous issue - no warranty.

 

The moral of the story is simple.  With warranty clocks that start ticking on the day of purchase, the price per GB continuing to decline, unraid now supporting dual parity, and with almost every online vendor offering 1-2 day shipping, I cannot think of a single reason to keep spare, rapidly depreciating drives around.  

Which is why I only keep a single spare of each size disk as backup.  When I swap out all drives of that size it becomes an offline backup drive like most of the drives of that size that I swapped out.

I keep only an ‘emergency’ spare that has reallocated sectors (which have proved stable across pre-clear cycles) and that I do not want long term in my array.   As I also have Dual parity and it is easy to get next day (or often even Same Day) delivery from Amazon I probably do not even need that!

5 hours ago, BobPhoenix said:

Which is why I only keep a single spare of each size disk as backup.  When I swap out all drives of that size it becomes an offline backup drive like most of the drives of that size that I swapped out.

I also make sure that my array has so much room that I can just empty a drive towards the rest of the array (unbalance helps in that).

 

This means that when a drive fails I move the contents of that failed drive (that is then emulated) towards other parts of the array. The failed drive that is beiing emulated is then only an "empty drive". I can just remove the drive from the array and reset parity, that gives me time to preclear a new drive thoroughly and put it in..

 

I also have dual parity. Works for me.

Yep that works great if you have the space.  Have done something like that myself on one of my servers.  Once I copied everything off the failing drive I swapped out my second parity drive to replace it and copied the data back.  When I can afford to get 2 6TB drives I will add the second parity back and keep the second as cold spare.