"Workflow Executor" node suggestion

[HaveF]

Hi, KNIME developer ,

I have a suggestion about the node “Workflow Executor”. This node is super useful, I like it. Here is my idea about its option Debugging: Show executing workflow segment

When we enable this option.

If the workflow fails to execute and this option is enabled, the metanode will remain in order to be able to identify the problem.

The most scenario works great, except some conditions, like there is no result! Sometimes, it is not right if the output is an empty table, so, we cannot get in the metanode to see what really happens. Besides, there are some conditions we want to see the metanode even it is execute successful.

So I suggest:

1. no matter what we enable this option or not, if the node fail, the metanode remains.
2. if we enable this option, the metanode always remains.

Thanks for your attention!

[rsrudd]

@HaveF

Could you provide workflow snippet example that is causing this issue? That will be easier to get to the root of what exactly you are hoping to see with this added functionality.

Thanks,
Ryan

[aworker]

Hi @HaveF

If I’m not mistaken, one can have the called sub-workflow opened in a second tab of the KNIME editor and when you call and execute it, one can see how its execution and where it fails if this happens. This is the behaviour I have seen so far and it sounds good enough for debugging as in any other open component.

Did you consider this solution ? Maybe I’m missing the point

Hope it helps.

Best
Ael

[HaveF]

@rsrudd Thanks for your attention. Here is my workflow:

The upper part is normal capture workflow. The lower part is call the above workflow.

And I checked the Debugging option.

If the workflow(Node 11) failed, there will be a metanode nearby which I can see what’s wrong.

If the workflow(Node 11) not fail but output a empty table, like the picture. Then I don’t know what happens in the workflow execution.

Did I express it clearly enough?

workflow_executor_options.knar (43.5 KB)

[HaveF]

Hi, @aworker

Thanks. But it seems we are not in the same thing. See my example workflow above.

[aworker]

Hi @HaveF

Thanks for clarifying and for posting your workflow.

Definitely I was missing the point because you are using the -Workflow Executor- node and I was thinking of the -Call Local Workflow- node which has the behaviour I mentioned. You are completely right about the behaviour of the former.

I was biased towards the -Call Local Workflow- node when reading your post because recently I posted the following solution:

Better alternative to CASE SWITCH (with 15 options)?

Hi @kowisoft Sorry I’m replying late to your posts and questions about the JSON implementation since it is only now I have access to a computer with KNIME. The good news is that you got the idea and managed to implement it eventually ! I had a look at your solution and it could be implemented in a simpler way just by using the -Call Local Workflow (Table Based)- and the -Container Input (Table)- nodes. I do not think one needs the JSON versions in your case although it is a very good e…

which is very similar to what you are doing but based on the -Call Local Workflow- node in case you would like to have a look at it as an alternative to call external workflows.

Best
Ael

[HaveF]

Interesting.

Thanks for your link. I tried. It seems needs more work, after all it is a tradition way

Let’s forward to a better integrated deployment solution. highly recommended

[rsrudd]

@HaveF Thank you for your further explanation and desire to see KNIME change for the better.

I have marked this down and sent this to the Development team for review.

Regards,
Ryan