Operator manual¶
Configure AntFlow's connected services, user authentication, and execution infrastructure. For the initial deployment, follow Installation with Helm.
Use the examples¶
Configure kubectl for the cluster you intend to administer and keep using that
context for the deployment procedures. Commands create or update resources in
that cluster; they do not configure the hosted demo.
Configuration blocks are templates. Domains under example.com, storage paths,
account names, and values in angle brackets must be replaced with your own
service settings. YAML, TOML, and dotenv examples do not expand shell variables.
Keep names consistent with the resources you create, and preserve placeholders
explicitly documented as runtime substitutions.
For Server API commands, first follow Access the API
to configure API_URL and TOKEN and verify access to your intended instance.
Launcher registration tokens are separate credentials, not user API tokens.
Deploy and expose AntFlow¶
- Installation with Helm: deploy AntFlow and verify local access.
- Service connectivity: outbound connections for browsers, servers, launchers, and workflow jobs.
- Expose AntFlow over HTTPS: public endpoints, Ingress, Gateway API, and TLS.
- Configure cross-origin access: API CORS and bucket CORS for browser requests.
Configure identity and repositories¶
- Authentication: OIDC clients, group claims, user secrets, and encryption keys.
- GitLab projects: repository configuration, discovery filters, and project publication.
Manage application data and results¶
- Configure the application database: PostgreSQL, persistent volumes, or an existing database service.
- Store and serve workflow results: S3 bucket credentials, object paths, proxied reads, and presigned downloads.
- Publish results to STAC: static catalogs and transactional STAC API publication.
- Configure caching: in-memory, Valkey, or Redis backends, cache lifetime, and security.
Run workflows¶
- Launchers: registration and deployment on Kubernetes, local hosts, or HPC systems.
- Execution policy: queue priorities, launcher capabilities, retries, and API sorting.
Customize the platform¶
- Theming and branding: logos, navigation bar, and light/dark colors.
- Integrations: optional AntFlow Spaces and AntFlow Trigger connections.
Configuration reference¶
- Helm chart reference: chart values and deployment topologies.