no you are right, that error is certainly not normal.
However, to me it sounds like the KNIME AP runs out of memory which is quite possible with the random forest nodes.
The reason for this is, that the random forest learner loads the complete input table into memory in order to allow for a fast learning process. If that error keeps popping up, you could try to give more memory to your KNIME AP but I fear there is currently no other solution if that does not succeed.