Kine With YugabyteDB

==========================

This fork adds support for YugabyteDB backend. The YugabyteDB version is based on the PostgreSQL backend with several optimizations relevant for distributed SQL databases.

To use the YugabyteDB backend, pass the yugabytedb as a driver name in your connection string. For instance, a complete connection string can look as follows:

yugabytedb://yugabyte:yugabyte@127.0.0.1:5433/yugabyte

A YugabyteDB cluster needs to be started with the ysql_sequence_cache_minval=1 TServer flag. By default, YugabyteDB creates sequences with CACHE=100 which means that one connection to the database can insert records with ids in the range from 1 to 100 while the other in the range from 101 to 200. This can lead to version mismatch errors on the Kubernetes end that depends on sequences that increment strictly by 1 (meaning the sequences need to be created with CACHE=1). Here is an example for Docker that start a YugabyteDB node using the yugabyted tool:

docker run -d --name yugabytedb_node1 --net custom-network \
  -p 15433:15433 -p 7001:7000 -p 9000:9000 -p 5433:5433 \
  -v ~/yb_docker_data/node1:/home/yugabyte/yb_data --restart unless-stopped \
  yugabytedb/yugabyte:latest \
  bin/yugabyted start --tserver_flags="ysql_sequence_cache_minval=1" \
  --base_dir=/home/yugabyte/yb_data --daemon=false

Kine is an etcdshim that translates etcd API to:

SQLite
Postgres
MySQL
YugabyteDB
NATS Jetstream

Features

Can be ran standalone so any k8s (not just K3s) can use Kine
Implements a subset of etcdAPI (not usable at all for general purpose etcd)
Translates etcdTX calls into the desired API (Create, Update, Delete)

See an example.

dmagda/kine-yugabytedb

Kine With YugabyteDB

Features