I run multiple servers in our cluster and there is always an AMI backup that runs before any backups are made. So there is constantly a redundant backup of the server.
I recently upgraded from Ubuntu 18 LTS to Ubuntu 20 LTS without issue, and had full redundancy in place in the event of an issue arising in production.
In addition, we use Github to manage all code, so there is a full backup of all code outside of the cluster.
In addition, we leverage RDS with multi-region redundancy so there is very little chance of losing any data.