The config datasource¶

Overview¶

The config datasource is a Congress datasource that let you access the definition of configuration options of Openstack services on the nodes of the cloud.

Its main purpose is to let administrator check the coherence of a deployment by defining constraints on the value of those options.

To retrieve the configuration files, the datasource uses an agent that must be configured on the various nodes.

Agent configuration options¶

Please refer to Configuration Options for the generation of a sample configuration file (etc/congress-agent.conf.sample).

The options most important to the agent are described below. They all appear under the [agent] section of the configuration file.

host

The name of the host as it will appear in the Congress tables.

version

This is the Openstack version used on the host. It can be used for rules that only apply to a given version of a service. Standard version names in lowercase should be used (e.g. ocata, pike).

services

This is a dictionary that associates to each service (use standard service names in lowercase), a description of its configuration file. The description is itself a dictionary. Keys are the paths to the configuration files and values are the path to the template definitions of the configuration files for oslo-config generator.

The agent also requires a valid configuration of the message bus as it uses it to communicate with the datasource driver.

An example of configuration file is the following:

[DEFAULT]
transport_url = rabbit://...

[cfg_validator]
host = compute-node-001
version = pike
services = nova: { /etc/nova/nova.conf:/opt/stack/nova/etc/nova/nova-config-generator.conf },neutron: { /etc/neutron/neutron.conf:/opt/stack/neutron/etc/oslo-config-generator/neutron.conf },congress: { /etc/congress/congress.conf:/opt/stack/congress/etc/congress-config-generator.conf }