Hi @Guglielmo ,
which version of KNIME Analytics Platform are you using now?
Could you provide us with the WFs without sensitive information so we could look into what is happening there?
I am not sure if you work on such a use case for the first time or you are already experienced with it. If it is your first time, you can maybe look into this WF to check if both WFs contain what’s required.
Best,
Daria
Hi @Darspir I’m working with version 4.7.7. I cannot provide the original WFs where I’m working as they contain many company information, in the next weeked I can prepare some similar workflows removing all sensitive information.
Frankly speaking I’m not experienced with this type of solution, actually I’m looking a way to transfer “automatically” the data from one workflow to onother and I find a solution on the KNIME Hub so I readapted this solution to my case.
I looked at the suggested workflow but it seems a bit different from what I need.
In my original post I omitted the called WF, see below:
Hi @Guglielmo ! thanks for reaching out to us. These nodes can indeed be tricky and, unfortunately, the error message here doesn’t help a lot.
Could you please tell us:
did you provide the name of the workflow explicitly or did you click “Browse” and selected the workflow? The second option is more reliable - please try that if you haven’t yet.
are your workflows in a LOCAL workspace or on KNIME Server or on KNIME Business Hub? If in LOCAL workspace, the Mountpoint Connector is a good option, if on Server or Hub, I’d recommend using KNIME Server Connector or Space Connector respectively. This is especially important for Business Hub because the folder structure is a bit different (teams → spaces → folders) and Space Connector takes care of that.
Finally, if you’re calling workflows on Business Hub, do you have a default execution context set up? This can be also a problem but using the Space Connector should solve this.
If you are looking for the way to transfer data (or any other objects) between KNIME workflows, there is an easier way to do it with Workflow Service nodes (Call Workflow Service, Workflow Service Input / Output). I created a small example here similar to your screenshots. These nodes are easier to use, you don’t need to transform data to and from JSON, they actually can transfer any objects for which KNIME has ports, and of course the data tables. They are also faster. The logic is similar to what you did - Input and Output nodes in the called workflow and Call … node in the calling workflow - but it would simplify your work a lot.
Here in this blog post you can find even more guidance on all the workflow invocation nodes.