Currently, the kea-python hook requires the user to provide the libpython shared object and the target Python. This works well, but it implicitly uses the operating system Python environment. To support isolated environments and allow for dependency management via venv, it would be useful to allow users to specify a Python virtual environment to be used instead.
Config Example
{
"library": "/lib/x86_64-linux-gnu/kea/hooks/libdhcp_python.so",
"parameters": {
"libpython": "libpython3.8.so.1",
"virtualenv": "/opt/myapp/",
"module": "kea.hooks"
}
}
Benefits
- Better isolation of dependencies
- Easier integration in multi-project environments
- Avoids polluting system Python, specially in moder systems
- Supports different runtime configurations for different KEA instances
Currently, the kea-python hook requires the user to provide the libpython shared object and the target Python. This works well, but it implicitly uses the operating system Python environment. To support isolated environments and allow for dependency management via venv, it would be useful to allow users to specify a Python virtual environment to be used instead.
Config Example
{ "library": "/lib/x86_64-linux-gnu/kea/hooks/libdhcp_python.so", "parameters": { "libpython": "libpython3.8.so.1", "virtualenv": "/opt/myapp/", "module": "kea.hooks" } }Benefits