TODO

- [DONE] Detect and Automatically recover from Quorum failure
  - Cron job that periodically checks status
  and if it had properly bootstrapped but currently
  there is no quorum, bootstrap the instance again.
    - Ensure that no two or more instances of
    this cron job is running at the same time.

- Backup zookeeper data, load it on recovery
  - Only applicable to fresh bootstraps (when ZK failed)

- [DONE] Remove terminated nodes on scaling down
- [DONE] Test scaling up and down Zookeeper to see if quorum and
majority scales as well.



- Test ansible playbook
- Test CF script
- Test zookeeper-utils
- Test with Kafka


- [DONE] README.md with instructions how to build and deploy the ami
  - To build:
    >> cd ami
    >> ./build.sh
  - To deploy:
    >> cd cf
    >> vi deploy.sh and update the --ami Parameter
    >> ./deploy.sh