Reference architecture for the OpenVox platform.

This is our [WIP] opinionated reference architecture for the OpenVox platform. This implements the following node types:

1. The primary OpenVox server:
   • OpenVox agent
   • OpenVox server
   • OpenVoxDB
   • Foreman web UI
     • Redis Rails cache
     • the preferred option of non-EOL vendor provided PostgreSQL
   • Regular PostgreSQL backups, including both Foreman and OpenvoxDB tables
   • Full two-way integration between OpenVox and Foreman
   • [todo] Performance tuning profiles
2. [todo] The secondary OpenVox compiler nodes:
   • OpenVox agent
   • OpenVox server
   • Codebase deployment configured
   • Central reporting
   • Performance tuning profiles
3. [todo] OpenVox agent nodes:
   • OpenVox agent
   • Optional serverless configuration
     • Codebase deployment configured
     • Central reporting

⚠️ This module is still under heavy development and the interface is expected to change drastically. Do not take configuration choices we've made with this module as recommendations (yet). See #limitations and don't use the module yet. Contributions and suggestions are very happily accepted.

Node Ownership

The server and compiler profiles are intended to have full ownership of a node. You may add on things like Foreman smart proxies or plugins such as the Hiera Data Manager, but don't also try to run your company website or anything on it.

Usage

Primary OpenVox Server

This is the most common use of this module. Simply include it on a node to configure a primary server with OpenVoxDB and a standard Foreman configuration.

include openvox_platform

You may also choose to disable Foreman or OpenVoxDB to get these standard stacks:

Foreman	OpenVoxDB	standard configuration
✅	✅	primary OpenVox server
⛔️	✅	no GUI primary OpenVox server
⛔️	⛔️	secondary OpenVox compiler node
✅	⛔️	no OpenVoxDB (not recommended)

There are several optional configuration parameters you may set. Please see the class documentation for more information.

Advanced usage

If you would like to customize the Foreman, OpenVoxDB, or PostgreSQL beyond the options provided, you may use the standard Hiera parameters for the underlying component modules. If you have a support plan then check with your provider for support options.

Secondary OpenVox Compilers

[todo] To configure a node as a secondary compiler, just classify it without enabling Foreman or OpenVoxDB. It will configure itself with a copy of the Puppet codebase and to send reports back to the primary server.

class { 'openvox_platform':
  foreman  => false,
  puppetdb => false,
}

OpenVox Agents

[todo] Use the openvox_platform::agent class to configure agents. You should include this on all nodes in your infrastructure. If you intend to run serverless then specify a control_repository and optionally an update frequency.

Limitations

This module is not ready for production use yet. You are welcome to kick the tires and poke & prod all you like, but please don't rely on it yet. It will be heavily refactored as we flesh out how the platform will work. Parameters will change and the way data & parameters are passed will change. Class names will be refactored and may or may not have a 1-1 translation. This is currently only regularly tested on x86 CentOS 9 until we set up proper pipeline testing.