My own attempt at building a distributed filesystem.
- Build nodes' APIs (not REST endpoints or anything yet, just functions that can be called) for IO + deal with duplicates/delays/misorderings
- Use Raft for replication
- Build client to figure out which nodes to contact and how to split up the file into fragments
- Change APIs to http endpoints and test using Kubernetes
- Set up nodes on the cloud
- Profit