The KNV project aims to be a learning tool for a basic key-value database with two twists:
- A single value may be listed under multiple indexes within a table
- Distributed Database
KNV expects to run with the following services:
- Etcd
The following are currently supported:
- gRPC-based multi-key-value store API
- local on-disk storage
- Cluster Membership
Up next are the following features:
- Write re-direction to shard write-leaders
- Kafka-backed storage
- Kafka-enabled storage replication
- Separate read instances
- Self-healing node/shard assignment