We have a Knime server running on AWS, using the latest images from AWS Marketplace.
Python nodes do not work - I get the following error:
2021-01-28 12:41:37,427 : ERROR : KNIME-Worker-64-Python Script (1⇒1) (deprecated) 0:1265:36 : f387cb8a-1f9f-4512-aaed-4a3d59370848 : Node : Python Script (1⇒1) (deprecated) : 0:1265:36 : Execute failed: Could not start Python kernel. Error during Python installation test: Could not start Python executable at the given location (no_conda_environment_selected/bin/python): Cannot run program “no_conda_environment_selected/bin/python”: error=2, No such file or directory
However my configuration is “standard”. What can be noted though:
I have a custom preferences.epf ( for my palladian license )
I updated the server to 4.11.3 and made a fresh executor install of version 4.3.0 ( using the “KNIME Server Executor” package from the download page ). I don’t know if the python nodes were working before the update.
I see here and on my server that a default setup is in place including conda and a default environment.
What is the missing piece to get the Knime server work with the preinstalled Conda environment?
Maybe some default preferences.epf lines?
In the meantime I got it working by adding the following lines to my preferences file: ( ie. /srv/knime-server/config/client-profiles/executor/executor.epf )
However it would have been a great time-saver if those three lines were there by default, or at least written somewhere in the doc!
( Sorry if it’s there and I didn’t find it! )
Thank you for this information. Everyone’s environment is a little bit different, and we have users who deploy on both Linux and Windows, so it may not be ideal to include it as a default.
The necessary documentation for enabling and using the preferences files in KS 4.11 can be found here [1]; however, I will forward your feedback about the possibility of including your recommendations in the documentation for users such as yourself who are seeking enlightenment about options.
Ha! Fair enough. But the main point is that we will communicate to the documentation team that it may be a good idea to include such options in documentation so that they are visible to those who need them, and to thank you for bringing it to our attention.
