A high availability MySQL server on AWS is about the same difficulty as on your own kubernetes instance (I've got a play one on one of those $100 N100 machines, got one with 16G mem). Then:

    helm repo add mariadb-operator https://mariadb-operator.github.io/mariadb-operator
    helm install mariadb-operator mariadb-operator/mariadb-operator
And then you can just provision MariaDB "kind", ie. you kubectl apply with something specifying database name, maximum memory, type of high availability (single primary or multimaster) and secret reference and there you go: new database, ready to be plugged into other pods.

Dont you need ECC in your db nodes?

N100 is my homelab, for playing. For instance I have a kubernetes cluster running KubeVirt, which runs 5 VMs, which ... have a kubernetes installation (so I have multiple worker nodes doing a "distributed filesystem" all of which is resharing disks from the same SSD). My production servers are generally older Xeons with ECC ram, which are also running kubernetes.

N100 supports DDR5 memory (although 1 channel) but I believe DDR5 has some error correction... May not be full ECC

amazing how nobody even know about ECC these days.

see so many series B+ companies running DB and storage without a care in the world.