Conda Environment Propagation with Call Workflow

Hi Everyone :slight_smile:

I have set up a master and worker workflow combination, where the master workflow calls the worker workflow using the -Call Workflow (Table based)- node in KNIME 4.7.8.

The worker workflow starts with a -Container Input (Table)- node and a -Container Input (variable)- node so that I can pass both data and variables from the master to the worker.

I would like to use the -Conda Environment Propagation- node in the master workflow and pass the environment as a variable to the worker workflow through the -Call Workflow (Table based)- node. However, the variable does not appear in the -Container Input (variable)- node when the workflow is called.

Please could anyone advise how to resolve this problem? Is it similar to the issue where Path variables are not passed to the worker workflow using the -Call Workflow (Table based)- node?

Thanks a lot

Hi @HeatherPikairos,

That’s interesting, I have never tried this myself.
But I think that the Container nodes have to transform - under the hood - the objects to JSON when passing it to the callee. This might cause problems with the CondaEnvironment Type variable. Have you had a try with Workflow Services Input and Outputs?



Hi @emilio_s

Thanks for the suggestion! I have just tested a simple workflow with the Workflow Services nodes and I’m able to pass the CondaEnvironment variable to the callee!

I will use these nodes in future workflows :slight_smile:

Best wishes


Just for clarification, This was done on the hub and not locally with Workflow service node right?

Hi @Daniel_Weikert

Actually I was just testing this locally and not on the hub


1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.