HAProxy Deployment

This repository holds both the spojenet.haproxy Ansible role (roles/haproxy/) and a ready-to-run playbook (playbook.yml) that deploys/reconfigures haproxy.spojenet.cz. It is designed to be self-contained: cloning this repo and running the playbook is enough to manage that HAProxy instance, no other repo required.

Production deployment runs automatically via Semaphore (semaphore.proxy.spojenet.cz): a push to the main branch triggers a webhook that re-runs playbook.yml against haproxy.spojenet.cz.

Repository layout

playbook.yml          # entry-point playbook (hosts: haproxy.spojenet.cz)
vars/haproxy.yml       # haproxy_backend_servers, certbot_admin_email, nodes
inventory/hosts        # static inventory for local/manual runs
requirements.yml       # external galaxy role dependencies (geerlingguy.certbot)
roles/haproxy/         # the spojenet.haproxy role itself

Local usage

ansible-galaxy install -r requirements.yml
ansible-playbook -i inventory/hosts playbook.yml
# or: make haproxy

Requirements

• Ansible 2.9 or higher.
• The target system should be a Linux-based operating system.
• Ensure the haproxy package is available in the system's package manager or provide a custom source.

Role Variables

The following variables can be customized to suit your needs:

• haproxy_global_config: A dictionary defining global HAProxy settings (e.g., log, maxconn, etc.).
• haproxy_defaults_config: A dictionary defining default HAProxy settings (e.g., timeout, retries, etc.).
• haproxy_frontends: A list of frontend configurations, including ports and ACLs.
• haproxy_backends: A list of backend configurations, including servers and load-balancing algorithms. Each backend server can include:
  • name: The name of the backend server.
  • address: The address of the backend server (e.g., 10.11.56.151:8080).
  • httpchk (optional): The HTTP path to check for health monitoring (e.g., /api/json).
  • expect (optional): The expected HTTP status code for health checks (e.g., 200).
  • ssl: Whether SSL is enabled for the backend server (true or false).
• haproxy_service_state: Desired state of the HAProxy service (started, stopped, etc.).

For a full list of variables, refer to defaults/main.yml and vars/main.yml.

Dependencies

This role does not have any external dependencies. However, ensure that the target system meets the requirements mentioned above.

Example Playbook

Here is an example of how to use this role:

- hosts: load_balancers
  become: yes
  roles:
    - role: spojenet.haproxy
      haproxy_global_config:
        log: "/dev/log local0"
        maxconn: 2048
      haproxy_frontends:
        - name: http_front
          bind: "*:80"
          default_backend: web_servers
      haproxy_backends:
        - name: web_servers
          servers:
            - name: jenkins.proxy.spojenet.cz
              address: 10.11.56.151:8080
              httpchk: /api/json
              expect: 200
              ssl: false
            - name: www.domain.com
              address: https://domain.com

License

BSD

Author Information

An optional section for the role authors to include contact information, or a website (HTML is not allowed).