10.8 Node Configuration

Go To Top

Return to Node Implementation Pattern

Each Node should load Configuration from approved Configuration files, environment variables, command-line arguments, or container settings. The Node profile should identify the required Configuration sources and values.

Node Configuration should define:

1. Node Identity.
2. Node Role.
3. DDS participant settings.
4. Topic names.
5. QoS profile references.
6. Logging settings.
7. Runtime mode or demonstration profile.
8. Required input and output locations, where applicable.
9. Command-handling settings, where applicable.
10. Status-reporting settings, where applicable.

Configuration should avoid hard-coded machine-specific assumptions. Nodes should not depend on local usernames, personal directories, hidden shell profiles, undocumented environment variables, or editor settings.

The team should distinguish Baseline Configuration from local overrides. Baseline Configuration should support repeatable execution. Local overrides should remain clearly marked and should not be entered into the controlled Baseline unless the team explicitly approves them.

A Node should validate its Configuration before completing startup. If Configuration validation fails, the Node should log the failure, publish a Failed status where possible, and exit with a meaningful error code.