The rs-server offers two deployment targets:

  • cluster
  • local

For each kind of deployment, here are the overall associated features.

Cluster mode

On the cluster mode, the rs-server is deployed as several Kubernetes PODs. A deployment with HELM chart is provided. On the cluster environment, rs-server can be called from a Prefect chain or directly from a Jupyter notebook. The rs-server running the cluster has a strong access control for each sensitive data access (CADIP, LTA, AUXIP, PRIP, and Catalog).

Here are CSC RS Python in a Nutshell features from the cluster.

Local mode

On local mode, rs-server, the station mockups, Prefect, Dask and a Jupyter server are deployed on the local machine with Docker Compose, without access control. It is used for development and tests: from the local machine, the end user can read and write the local rs-server catalog, and run the Prefect flows locally.

See the RS-Python documentation for the details of both modes.

The differences between both modes:

Local mode and cluster mode