Not able to see visualization when deployed to Knime server

I have the following workflow where I have a bar chart visualization.

I am able to deploy it to a Knime server and execute it successfully but am not able to see the Bar plot once the execution completes succesfully. I am able to see the Green Success message but am not able to see the Bar plot on the server. What am I missing here ?

Could you try to export your graphic? I think if you open a workflow directly from the server some results are not displayed.

Export could be something like the later part of this workflow.

Hi @sramesh,

Can you please describe how you are trying to view the visualization? Via WebPortal, or from opening the remote job?

Cheers,
Roland

I have deployed the Knime report on server and had run the job and was trying to view it from Knime web portal.

Like once the job is completed I was thinking it would show the visualization.

I had also tried having a image writer port in the work flow and had published it but am still not seeing any visualization after deploying and running the job on server

Hi @sramesh,

does capture you Bar Chart node into an Component helps? Right click the node -> Create Component …

gr
Hans

2 Likes

Tried encapsulating the bar chart into component, still not working.

Does it work on your local machine? Right click the component: Interactive View: results

1 Like

I am able to view it on the local machine, I am also able to export the image and save it locally. But I am not able to view it on the Server. I am able to deploy it to the server and even run it but am not able to view the graph once the workflow executes successfully.

Hi @sramesh

can you try to download this workflow from the hub https://hub.knime.com/knime/spaces/Examples/latest/09_Enterprise/02_WebPortal/07_Visualize_Scatterplot_on_file

Afterwards upload it to the server and test if this is working for you.

Best, Iris

I was able to download the workflow and it looked like I had to import a csv file to visualize a scatter plot and so I had imported a csv and was able to run the workflow on the local machine and get a scatter plot.

I had then deployed the workflow and ran it on the server and got the same result, it executed successfully on the server but does not display any plot.

I had tried to deploy other example visualizations on the server but am not able to produce a visualization on the server, Can you please let me know if there is any setting on the server that I need to check to get the plots ? like what am I missing here.

Can you please confirm whether the workflow was already executed when you started it on WebPortal? You need to save the workflow in a reset state in order for it to show visualizations on the WebPortal.

2 Likes

Thank you for letting me know that. I did not reset before deploying to the server which I did now and was able to view the visualization for the example scatterplot workflow.

However , I am trying to replicate the same workflow for a POC and am not able to get it to display the visualization. I have a file upload node which feeds into csv reader and does a scatter plot

I am able to run the workflow on the laptop, I took care to reset the workflow and then deploy to server , Then tried executing the workflow , am able to see a green success message but does not prompt for fileupload and does not display visualization.

I even took out the database reader node to make it closer to the scatter plot example but am still not able to generate the visualization. Any pointers here please ?

Hi @sramesh -

Looking at the outline pane, I don’t see any component inputs or outputs. So I’m guessing you haven’t wrapped the File Upload and viewer nodes into components, which is necessary for display in the Webportal.

3 Likes

That worked!. Thank you, I am able to upload a file and visualize the plots on Knime server.

1 Like

Hi guys,
I’m experiencing the same problem of “empty” WebPortal visualizations, while there are supposed to be the “many” included in one component.
Luckily the suggestion of @RolandBurger still works:

Anyway, I think it would be (much) better if KNIME AP solve this problem at the root, during the deploy process, at least adding a WARNING message for alerting AP users about this (let me call it) anomaly in WebPortal behavior.
Thank you in advance

PB

Hi @paobar61,

This is actually the expected behavior, and not at all an anomaly :wink: The reasoning behind this is that we want to allow the user to deploy partially executed workflows, and then only resume execution from where the workflow was saved. E.g., you can upload a workflow in which data has already been read and processed, so that on the WebPortal, only the visualization has to be processed. This can significantly speed up the time it takes for the visualizations to be shown.

Obviously, this has the drawback that a fully executed workflow will not show anything. Still, this is hard to catch upon uploading the workflow, since at that point, we don’t actually know whether the workflow will run on the WebPortal or be called from the client / via REST. If in doubt, the “Deploy to Server” dialog does offer the option to reset before upload, which is a reasonable thing to do unless you have good reasons why you want to upload the workflow in executed state.

Cheers,
Roland

Thank you for your answer @RolandBurger,

I agree with you that, in some cases, could be nice to deploy a partially executed wf so that only the visualization has to be processed, but the problem I’ve experienced and notified is that if you deploy a wf “without” before resetting it, the WebPortal will not show any visualization at all, and I hope you’ll agree with me that this is not an expected behavior, won’t you?
Regards

PB

Hi @paobar61,

No, it is actually the expected behavior. I fully agree that this is not ideal, but it is expected :wink:

I have raised this with our developers. There is currently a lot happening in terms of WebPortal development, so I hope we can come up with a more reasonable solution for how to catch fully executed workflows.

Cheers,
Roland

1 Like

Thank you very much @RolandBurger, let’s hope :wink: