Hello .
I have a very strange workflow error which is hard to debug due to nearly no logging output. Also the workflow runs smooth in the GUI. And I cannot share the workflow file here with you - sorry.
Setup: KNIME AP 4.5.3 in batch mode on Windows. The batch file itself has several workflows it executes. They are all located on the same network share and running with the same startup parameters. As 9 of 10 are running this way, I want to exclude the setup and configuration as error source.
The KNIME “.log”-File does not contain any relevant information for the batch run. As far as I can see it, the log only contains outputs from the UI.
For debugging, I already restarted the machine to make sure not having a “hanging” KNIME process in the background (also checking this in the task manager). I tried the batch with different combinations of “-consoleLog”, “-noexit”, “-reset” (always), “-nosave”. The console log didn’t output something else than I can see in the GUI where the workflow is running.
As addition: The workflow writes a protocol to a database as the very last step, equal if there is workload or not. The protocol is written before error 4 is casted. So the error is casted at the very end of the workflow, maybe even in kind of post-workflow-operations, e.g. cleanup or kind of this.
Did anyone here had a similair issue with the execution of KNIME workflows yet? And in best case also found a solution?
Thanks a lot!