Question for the nerds: how do enterprises and critical systems etc stop this happening? I guess it makes sense that if you have 4 drives made at the same time, being used in the same way, they will fail at the same time or very close to each other. Making RAID more risky?
In certain critical scenarios, IT/facilities/whoever else has 2N+1 redundancy for everything, and it's commissioned to verify that system failures would be independent.
I'm not Josh, so I don't know exactly what he's working with, but if the chance of a single failure of a device is 1/100 every day (failures is actually quite normal for a device running 24/7), then having four
independent failures in one day, over the lifetime of a 10 year website would be at least 1 in 30 thousand (devices tend to fail at similar times, so it could be lower, maybe 1 in 10 thousand, or even 1 in 100), and with multiple things that could go wrong, it should not be surprising.
Big enterprises lower those odds by having multiple datacenters, made using different methods, in different storage areas (Google Cloud/Microsoft OneDrive is the most frequent backup for smaller enterprises), for national companies: on different power grids, and for huge conglomerates like Microsoft and Google: on different continents. Josh owns much of his equipment, and the total failure of the power grid in a single country could potentially take this website down. Then, in each datacenter, all the servers are independent, and literally every other piece of equipment, from transformers to HVAC equipment is 2N+1 redundant as well.
This is not such a big deal for Kiwifarms, which doesn't operate with a million of dollars of revenue every single hour (imagine how much money Google would lose if AdSense went down for an hour), but huge enterprises will build a huge fucking datacenter in another country and spend millions of dollars.
