README.md

Ansible modules

This is a collection of modules written internally at mnubo for use with our infrastructure stack. In the next few lines we'll highlight what's available and how to use them. We've included a requirements file for our modules in the library directory.

k8s_secret

A module to manage kubernetes secrets. Pretty useful when using private registries so Kubernetes can pull volumes by itself. Basic example:

  k8s_secret:
    config:
        api_host: "localhost:8888"
        namespace: "default"
    name: "my_registry"
    definition:
      secret_type: "dockercfg"
      data: '{"private-registry.example.com":{"auth":"erHyHVWyTgRbCLpKgXPdXYFxKNzEZGhRfGvcQ4o9c4pjvTchmk", "email":"[email protected]"}}'
    state: "present"

See module code for more documentation.

k8s_pod

A module to manage kubernetes pods. Basic example:

- name: Create a simple pod
  k8s_pod:
  config:
      api_host: "localhost:8888"
      namespace: "default"
      pull_secret: "my_registry"
  name: "myweb"
  spec:
      containers:
          - name: "myweb"
            image: "httpd:2.4"
            ports:
              - containerPort: 80
                name: "http"
                protocol: "TCP"
            env:
              MY_VAR: "some_value"
            volumes:
              - name: "my_volume"
                mount_path: "/data/html"
                read_only: False
      volumes:
          - type: "host"
            name: "my_volume"
            path: "/data/myweb/html"
  state: "present"

See module code for more documentation.

Notes: We need to improve error management at the kubernetes-py layer and bubble-up the errors to an ansible output variable.

k8s_rc

A module to manage kubernetes replication controllers. If anything changed, it will perform a rolling restart of the replication controller. Basic example:

k8s_rc:
  config:
      api_host: "localhost:8888"
      namespace: "default"
      pull_secret: "my_registry"
  name: "myweb"
  pod:
      containers:
          - name: "myweb"
            image: "httpd:2.4"
            ports:
              - containerPort: 80
                name: "http"
                protocol: "TCP"
            env:
              MY_VAR: "some_value"
            volumes:
              - name: "my_volume"
                mount_path: "/data/html"
                read_only: False
      volumes:
          - type: "host"
            name: "my_volume"
            path: "/data/myweb/html"
  replicas: 2
  state: "present"

See module code for more documentation.

Notes: Rolling restarts have default variables. Please see the module documention in the code for more information. We need to improve error management at the kubernetes-py layer and bubble-up the errors to an ansible output variable.

k8s_service

A module to manage kubernetes services. Here's a basic example:

  k8s_service:
    config:
        api_host: "localhost:8888"
        namespace: "default"
    name: "myweb"
    definition:
      service_type: "NodePort"
      ports:
        - port: 80
          target_port: "http"
          protocol: "TCP"
          node_port: 8030
      selector:
        name: "myweb"
      cluster_ip: "10.100.200.10"
    state: "present"

See module code for more documentation.

Notes: The module is set to retry for a while. We need to improve error management at the kubernetes-py layer and bubble-up the errors to an ansible output variable.

chronos_job

A module to manage jobs to through chronos/mesos.

Here is a basic example:

  chronos_job:
    chronos_hosts: "{{ groups['mesos-masters'] }}"
    port: 4400
    name: "sample-chronos-job"
    state: created
    start_time: "2015-11-11T00:00:00Z"
    run_interval: "P1D"
    epsilon: "PT6H"
    cpus: 1.0
    mem: 1024
    docker_image: "{{ docker_registry }}/sample-chronos-job:{{ sample_chronos_job_version }}"
    env: ["ENV={{ env_id }}","JAVA_OPTS=-Xmx512m -Xms512m"]
    owner_email: "{{ chronos_job_alert_email }}"
    uris: "{{ chronos_docker_cred_uri }}"
  run_once: true
  delegate_to: "{{ groups['mesos-masters'][0] }}"

See module code for more documentation, testing and implementation TODO.