StarlingX Plugin

The StarlingX plugin enables you to discover StarlingX systems and manage resources on StarlingX systems.

Start out by providing the credentials for an existing StarlingX system. The install workflow will gather information from the system, for example type and location.

To discover subclouds, execute the discover_and_deploy workflow over a Controller deployment. This workflow performs the following tasks:

• Discover all subclouds that are managed by the system controller (parent).
• Create a deployment group.
• Create a deployment for each discovered subcloud. Add the subcloud deployment to the deployment group.
• Execute batch install on the deployment group.

NOTE: The StarlingX Plugin is designed to work with the StarlingX blueprint, which is located in the examples directory of the plugin repository.

Requirements

• Python versions:
  • 3.6.x

Labels

The following are possible for a StarlingX system deployment:

• csys-obj-type
• csys-env-type
• csys-obj-parent
• csys-location-name
• csys-location-lat
• csys-location-long
• csys-wrcp-services

Sites

During installation the plugin requests a system’s location (latitude and longitude) from the StarlingX API, and creates a Site in Cloudify. You can see sites’ location on the map on the Dashboard screen.

Node Types

cloudify.nodes.starlingx.WRCP

This node represents a StarlingX System. A system can be a System Controller, Standalone system, or a Subcloud.

Properties

• use_external_resource: Always true. This parameters indicates that the resource already exists.

• client_config: A dictionary containing the following keys:

  • auth_url: The StarlingX system endpoint, including the protocol, IP, port, and path, e.g. https://172.20.0.1:5000/v3. IPv6 systems’ auth urls, should be provided with the IP surrounded by square brackets, e.g. https://[2001:0db8:85a3:0000:0000:8a2e:0370:7334]:5000/v3.
  • username: The StarlingX username.
  • api_key: The StarlingX password.
  • project_name: The client project name, default: admin.
  • user_domain_name: The client user domain name, default: Default.
  • project_domain_name: The client project domain name, default: Default.
  • region_name: The region name, either RegionOne for a system controller, or the subcloud name in the case that the system is a subcloud.
  • insecure: Whether to ignore certification validation when using https as the protocol..
  • cacert: The content of a Certificate Authority when using https as the protocol.

• resource_config: Parameters that describe the system in the StarlingX API. Currently, there is no need to provide these parameters.

Runtime Properties:

• subclouds: A dictionary of subclouds if the node is a system controller that has subclouds.

  • A key containing a number representing the subcloud number:
    • external_id: The number representing the subcloud number.
    • name: The subcloud name.
    • description: The subcloud descriptions.
    • location: The subcloud location, for example an address.
    • group_id: The subcloud’s group ID.
    • group_name: The subcloud’s group name.
    • oam_floating_ip: The IP of the subcloud.
    • management_state: The subcloud’s management state.

• resource_config:

  • external_id: The system’s ID.
  • name: The system’s name.
  • description: The system’s description.
  • location: The system’s location, for example an address.
  • system_type: The system type, for example, “all-in-one”.
  • system_mode: The system mode, for example, “simplex”.
  • region_name: The region name, for example “RegionOne”, or if a subcloud, the subcloud name.
  • latitude: The latitude of the system.
  • longitude: The longitude of the system.
  • distributed_cloud_role: The distributed_cloud_role, for example “subcloud”.

• hosts:

  • A key containing an UUID representing the host ID.
    • hostname: The host’s hostname.
    • personality: The host’s personality, for example “controller”.
    • capabilities: The host’s capabilities.
    • subfunctions: The host’s subfunctions.

• kube_clusters: The API response for the system’s kube_cluster object.

• k8s_cluster_name: If the system has a Kubernetes cluster, the kube cluster’s name.

• k8s_ip: The IP of the Kubernetes cluster, if the system has one.

• k8s_service_account_token: The service account token, if the system has a Kubernetes Cluster.

• k8s_cacert: The Kubernetes Cluster certificate authority if the system has a Kubernetes Cluster.

• k8s_admin_client_cert: The Kubernetes Cluster client certificate if the system has a Kubernetes Cluster.

• k8s_admin_client_key: The Kubernetes Cluster client key if the system has a Kubernetes Cluster.

• openstack_ip: The system’s Openstack IP if it hosts an Openstack system.

• openstack_key: The system’s Openstack key if it hosts an Openstack system.

Workflows

The following workflows are available for the in StarlingX plugin:

• Discover Subclouds
• Discover and deploy

1. Discover Subclouds

Objective: Discover subclouds of starlingx controllers.

Summary

For a single controller a hint is used or for all nodes in the deployment discover subclouds is used. We get the controller objects, then search those for subclouds related to that controller. The controller node instance runtime properties are then updated with a dict of related subclouds info.

2. Discover and deploy

Objective:

Discovers and deploys subclouds.

Summary:

Checks if there are discovered subclouds using the “Discover Subclouds” workflow, then gets the controller node instance, and uses the parent deployment’s capabilities to get auth data.

Checks if the deployment_id passed from the workflow is valid and if there are 2 subclouds or more in the controller node instance runtime properties. If not, raises an Exception.

Then, for each of the subclouds, checks if the subcloud has already been discovered and deployed. If not, checks if the address of the subcloud is ipv4 or ipv6, formats and dict of “inputs”, adds the labels, adds the id of each of the deployments to a list, as the inputs and labels.

To complete, calls deploy_subclouds with the data gathered in the previous step.

Stories

The StarlingX plugin supports the following stories:

• Enrolling a StarlingX System Controller.
• Discovering the subclouds of a StarlingX System Controller, and automatically deploying them.
• Manually enrolling a subcloud of a StarlingX System Controller.

NOTE: Manual enrollment requires you to assign the csys-obj-parent label.

Prerequisites

All stories require that you provide a secrets for the following values:

• username: The default secret name is starlingx_username. This is the username of the StarlingX system.
• api_key: The default secret name is starlingx_api_key. The is the password to your StarlingX system.
• cacert: The default secret name is starlingx_cacert. This is the content of the certificate authority file.

You are free to change the names of the secrets, however, you must provide the secret names in the deployment inputs when enrolling a new system.

• Enroll StarlingX System Controller: You will need the auth_url of the controller, for example, https://123.123.123.123:5000/v3.
• Enroll StarlingX Subcloud: You will need the auth_url of the controller, for example, https://123.123.123.123:5000/v3. You will also need the subcloud name, and provide it as the region_name input.
• Discover Subclouds - execute the discover_and_deploy workflow. Provide no parameters.

Tests and Set Up

Running the tests:

In the terminal, run the following commands:

export HTTP_PROXY=http://147.11.252.42:9090
export HTTPS_PROXY=http://147.11.252.42:9090
tox

Running tests without tox

To run tests directly with pytest, run the following:

python3.6 -m venv venv
source venv/bin/activate
pip install --upgrade pip
pip install -r test-requirements.txt
pip install -r dev-requirements.txt


python -m pytest

or
python -m pytest path/to/test/file

Build wagon with docker helper

sudo docker run -v "$(pwd):/packaging" cloudifyplatform/cloudify-centos-7-py3-wagon-builder

or

sudo docker run -e HTTP_PROXY="http://147.11.252.42:9090" -e HTTPS_PROXY="http://147.11.252.42:9090" -v "$(pwd):/packaging" cloudifyplatform/cloudify-centos-7-py3-wagon-builder

Validate wagon

virtualenv env3 -p /usr/bin/python3
source env3/bin/activate

pip install wagon[venv]
wagon validate ./generated_wagon_file.wgn
wagon install ./generated_wagon_file.wgn

ls env3/lib/python3.6/site-packages/ | grep conductor_wrcp_plugin
deactivate