User guide

Python consumers of data and service discovery information should use the Repertoire client, rubin.repertoire. This library is available from PyPI and can be declared as a dependency or installed with pip in the normal way:

pip install rubin-repertoire

Consumers in other languages, such as JavaScript, can use the Repertoire server API directly.

Python client

• Creating a Repertoire client
  • Phalanx configuration for Repertoire
  • Use the FastAPI dependency
  • Manually create a client
    • Override the base URL
    • Provide an HTTPX client
  • Next steps
• Finding services
  • Service types
  • Service APIs
    • REST API versions
  • Butler configuration
  • Next steps
• Listing datasets
  • Next steps
• Getting InfluxDB connection information
  • Listing available databases
  • Getting connection information
  • Getting credentials
• Listing applications
  • Next steps
• Discovery data for Nublado containers
  • Data format
  • Generating the data
• Testing users of the Repertoire client
  • Prerequisites
  • Mocking service discovery results

Server

• Repertoire server API