Check out the links below.
Its common for drives in a RAID to fail. RAID was invented to save people from losing all of their data if a drive fails. So the whole system is built from the fact that it fails. Also the whole system kind of breaks the system further. RAID trys to solve the problem of bad drives, and it does, but at the same time drives get bigger thus break more commonly (failure per byte isnt getting that much better according to articles below) and we access them more often (causing more strain/stress thus failure increases). RAID causes extra writes and reads that other wise wouldnt be there (writing to parity/mirror data, thus causing more pain on the drives). Now that sounds like RAID should not be used, however that is not the case. The benefits of RAID out weigh the cons of RAID by miles. First off the rate at which drives fail in RAID5 or RAID6 is something we can avoid (depending on the size of your data and how much you use it), also if your using things like RAID10,50,60 then you can forget about any negatives about RAID (aside from cost). Oh another important factor to failing drives, is that you could simply have a drive model that fails often (check out links below for models of drives that dont fail often — oh dont forget the links dont mention every drive model for those manufacturers it talks about – i.e. Seagate & Western Digital are shown to be worse off than Hitachi, but in reality it all depends on the model – backblaze didnt test every model from every vendor)
Anyhow check out these articles that explain what Im trying to explain:
Best Drives & Worst Drives:
So if RAID5 and RAID6 are going to become obsolete, what can we use? Well maybe in the future RAID7 comes out (3 drive parity, currently ZFS does that RAIDZ3), but also as mentioned before you can use RAID10,50,60.