Ansible Execution Environments

Configuration Framework Service (CFS) sessions are comprised of a single Kubernetes pod with several containers. The inventory and git-clone containers run first, and a teardown container runs last (if the session is running an image customization).

The container that runs the Ansible code cloned from the Git repositories in the configuration layers is the Ansible Execution Environments (AEE). The AEE is provided as a SLES-based docker image, which includes Ansible version 2.9.11 installed using Python 3.6. In addition to the base Ansible installation, CFS also includes several Ansible modules and plug-ins that are required for CFS and Ansible to work properly on the system.

The following modules and plug-ins are available:

  • cfs_aggregator.py Callback Plug-in

    This callback plug-in is included to relay playbook execution results back to CFS for the purpose of tracking session status and component state.

    WARNING: This plug-in is required for CFS to function properly and must not be removed from the ansible.cfg file.

  • cfs_linear and cfs_free Strategy Plug-ins

    CFS provides two strategy plug-ins, cfs_linear and cfs_free, which should be used in place of the stock Ansible free and linear playbook execution strategies.

    For more information about Ansible strategies, see the external Ansible playbook strategies documentation.

  • shasta_s3_cred.py Module

    This module is provided to obtain access to S3 credentials stored in Kubernetes secrets in the cluster, specifically secrets with names such as <service\>-s3-credentials.

    An example of using this module is as follows:

    - name: Retrieve credentials from abc-s3-credentials k8s secret
  shasta_s3_creds:
    k8s_secret: abc-s3-credentials
    k8s_namespace: services
  register: creds
  no_log: true

    The access key is available at \{\{ creds.access\_key \}\} and the secret key is at \{\{ creds.secret\_key \}\}.