For example, I wanted to know if I can make 1 Kubernetes cluster span multi region, multi cloud

… For example, I wanted to know if I can make 1 Kubernetes cluster span multi region, multi cloud. So I slapped TailScale for networking, replaced etcd with multi-region PostgreSQL Aurora, and span the Kubelets between my Raspberry Pi, Digital Ocean VMs, and AWS EC2. And then as the "customer app", I run CockroachDB, rqlite, and tiDB on it (one at a time, I don't want to burn money for this). It was janky, zero SLA, I likely mis-tuned all the databases, and cost a bit of money :(. But it totally worked, all the db nodes can discover each other. I was satisfied and that's enough. Another example would be writing my own Raft-backed database, similar to the ToyDB Rust project posted here. Is the DB useful? No. But it is so much fun and I learned a lot. My AWS S3 bucket and private Git repo are a graveyard full of toy projects. My own Dropbox clone, Pinterest clone, Delicious clone, subset of Google Maps clone, etc. etc. are all RIP in there.