Nodes rolling restart

[Nuttymoon]

Linked issues

• Fixes Implement rolling restart mechanism #78

Changes

• node role
  • Add the stop and restart playbooks
  • Add the health-checks playbook that can take 4 parameters and sets 3 vars:

    Parameters:
    - wait_bootstrapped: Wait for the node to be bootstrapped on all chains (default: false)
    - wait_healthy: Wait for the node to be healthy on all Subnets (default: false)
    - avalanchego_health_checks_retries: Number of retries for health checks (default: 60)
    - avalanchego_health_checks_delay: Delay between retries for health checks in seconds (default: 10)
    Outputs:
    - avalanchego_state: State of the avalanchego.service (running/stopped/failed/unknown)
    - avalanchego_is_bootstrapped: Dictionary of bootstrap status on each chain (true/false)
    - avalanchego_is_healthy: Dictionary of healthy status on each Subnet (true/false)
    

• Add the rolling_restart playbook that makes use of health-checks with wait_bootstrapped and wait_healthy set to true

Additional comments

Note that it might not always make sense to wait for a node to be bootstrapped/healthy on all chains/Subnets.