Yep. Your example is the worst case scenario for a stripe set though. Other levels are a little different. And as you said, enterprise models make a difference for this reason.one drive say 10,000 hrs MTBF
two drives in a raid 10,000 MTBF each give a bathtub curve of 5,000 hrs MTBF
The data's spread out over multiple members (mechanical drives). So you've now more single points of failure. That's why they added redundancy to RAID (save a stripe set, which is the odd-ball of the levels, as it has no redundancy at all).Why does a Raid reduce the MTBF?
No, you don't need to get new drives. Those are enterprise grade units, and will last longer than 2 years. Mechanically speaking, 5 is closer to the reality, assuming you don't out-grow them in terms of capacity first.I have a 8 disk Raid of 500GB WD RE3 running 24/7 for about 2 years now. Should I consider swapping the disks with new ones?
I really want to keep that system as long as possible, as it is one of my backup devices.
You can mitigate issues by keeping the sets smaller. So dont' go and try a 24 disk stripe set with mechanical drives for example.