> A good critique for example is OVH lost a lot of customer data due to a fire. Where was the redundancy? That would make me think twice before switching to OVH.

I lost a VPS in that fire, but I was up and running a few hours later with a new VPS at a different OVH location.

Not to deflect blame away from OVH and their large screw up, but we should never rely only on the redundancy of the hosting provider. Even on AWS, I wouldn't trust them to not lose my data if one of their datacenters burns down.

At the time I was making regular backups to two different providers with servers somewhere else. When I noticed that it was serious, I ordered a new VPS and restored everything. If OVH itself went down, I could have used Scaleway, Hetzner, Contabo, etc.