The config key is where you can define the dynamic values that end customers need to configure before they can use your application. It might include things like:
- External Database connection details and credentials
- Other internal integrations settings like SMTP auth or API keys
- Tunable settings for your application like “number of worker processes” or “log level”
In Replicated Ship, a simple config section with one item might look like
config: v1: - name: "database_info" title: "Database Info" items: - name: pg_connstring title: "Connection string for the PostgreSQL server" type: text
The configuration options API is identical to that used for applications managed by Replicated’s scheduler suite, and the documentation can be found at Config Screen YAML.
Config values, but not defaults, will be saved at
.ship/state.json when assets are generated. On subsequent runs of ship, these saved values will be used instead of values within your config when preparing the initial settings screen for the user. If an item has
readonly: true set, then values from the saved state will not be used for that item. This allows users to have their existing settings prepopulated when rerunning ship.