Batch Executor ERROR KNIME-Worker-1 Node Execute failed: ("NullPointerException"): null

Hi Guys
I am facing this issue of nullpointerexception with very simple workflows, these were all working lately very smoothly.
have attached the two example flows that I used and the error screenshots.
Can anyone help what went wrong all of a sudden. I even had some very complicated workflows integrated with hive and spark dataframes that were working till yesterday but now none of those work and same error.
Thanks in advance.

Uploading: _generate_aMiniReport.knwf…


did you make any changes to your KNIME AP installation since yesterday, e.g., updating it?

Hi Mark
No, no changes made, my version is 4.0.1. interestingly the other installation ina separate machine is a 4.0.2 where also these batches are not working.


just for clarification. You didn’t make any changes to your AP, the workflow, or anything related to the workflow you are executing and everything worked just fine a couple of days ago, but now it is broken?

It is positive, am attaching the hs_err_pid2784.log (81.7 KB)
log file that was created every time I ran the batch. Also the batch script is shared below fyi.

“C:\Program Files\KNIME\knime.exe” -consoleLog -nosplash -nosave -reset -noexit -application org.knime.product.KNIME_BATCH_APPLICATION -workflowDir=“C:/Users/Administrator/knime-workspace/Reports/Generate Daily Sales Report Everyday”

Anything change with respect to the display (have you altered the Windows theme? are you now using multiple monitors? …) ?

The log hints at a problem with the native Java libraries fetching UI related assets.

Nothing as such happened Dear Quaeler, rather a few hours ago, I reinstalled Win 10 and tried running and same error showed up saying NullPointException.
Do you recommend some basic machine settings to be followed while running batch? I checked knime.ini and found 4033MB as default for heap space mgt. Pl suggest what should I check.

I have nothing off the top of my head - what is odd is that the NPE shown in the screenshot is in code that has nothing really to do with the stack trace shown in the .log file you attached above…

Should I send u both the workflows so u can check if possible?if yes, can u perhaps share ur email id?

Hi Guys
I did make the mistake of not putting the two dots before the wf name in Call local wokflow node but that was not the actual reason my workflows were not executing in batch, it was aparently the "SPARK Groupby node that has the problem, though my setup includes spark jobserver v2 which should have supported v2.43 jobs and cloudera spark v2 but its not as it seems. Anyway, I then wrote a spark query to replace the groupby node and it worked like a charm.
I think I am falling for KNIME big time.


Happy to hear that you were able to solve you problem!


1 Like

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