I have pretty much everything automated nicely. In the worst case scenario I can migrate to a totally different cloud provider and be up and running again in le

I have pretty much everything automated nicely. In the worst case scenario I can migrate to a totally different cloud provider and be up and running again in less than an hour. For the database I use an HA Postgres cluster so it can tolerate the temporary loss of one or two nodes. My dataset is still tiny so recovering is super quick.