## Enhancements

- `cluster-status` now shows which node a follower is following
- It is now possible to clean up a partially added node with `chef-backend-ctl remove-node`.
- Stability improvements in leader election

## Bug Fixes

- Ensure that elasticsearch uses the correct directory
- Do not create recovery.conf on leader
- Gracefully handle failover while syncing.
- Retry etcd connections when chef-backend-ctl is waiting for leader.
- in multi-IP hosts, only test the specified IP and not all.
- Do not overwrite existing configuration
- Other: Misc code cleanups