Sign up to take part
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
I have a custom plugin with some shared code in the "python-lib" directory that I'd like to import and use in other places.
Following this link - https://dss.grc-mlops.com/admin/code-envs/design/python/plugin_scaler-engine_managed/
The following code works in a regular notebook running on DSS. However, if I change the kernel to one that runs in a container - it results in an ImportError
Code;
import dataiku
plugin_id = "my_custom_plugin"
package_name = "foo"
dataiku.use_plugin_libs(plugin_id)
module = dataiku.import_from_plugin(plugin_id, package_name)
Error when running in a container:
ImportError: No module named 'foo'
Hi @AsishM
Answer can be found in https://doc.dataiku.com/dss/latest/containers/concepts.html#limitations,
"In code recipes and notebooks, using libraries from plugins is not supported in containers. For example, dataiku.use_plugin_libs and dataiku.import_from_plugin will not work."
Cheers,
Matthieu
Hi @AsishM
Answer can be found in https://doc.dataiku.com/dss/latest/containers/concepts.html#limitations,
"In code recipes and notebooks, using libraries from plugins is not supported in containers. For example, dataiku.use_plugin_libs and dataiku.import_from_plugin will not work."
Cheers,
Matthieu
Thank you! I didn't check the containers documentation. I imagine adding a note in the code recipe page also wouldn't be a bad idea