Specifying a custom Python interpreter

Cloudera Data Visualization includes an embedded Python interpreter by default. However, you can configure the system to use the OS-provided Python or a custom Python build instead. This provides greater flexibility and can improve performance when compatible Python versions are available on your system.

To use a non-embedded Python interpreter:
1. In Cloudera Manager, go to Clusters and select the Dataviz service.
2. Click the Configuration tab.
3. Set the following environment variable in the Dataviz Service Environment Advanced Configuration Snippet (Safety Valve) configuration to disable the embedded Python.
  DATAVIZ_USE_EMBEDDED_PYTHON=false
  
  This enables Cloudera Data Visualization to use the system’s default, OS-provided Python interpreter.
4. Restart the Dataviz service to apply the changes.
Optional: To use a custom Python build:

If you want to use a specific Python installation (for example one you compiled from source or installed with pyenv), follow these additional steps:
1. Set the PYTHONHOME environment variable to point to the root directory of the Python installation you want to use.
2. Configure the library paths by making sure that the following directories are accessible:
  - python/lib
  - sqlite/lib (only if you compiled SQLite with this Python build)
  You can do this by:
  - Adding them to the LD_LIBRARY_PATH environment variable, or
  - Registering them with ldconfig if you are using dynamically linked executables.