Simple Streaming: run single node in parallel


there is a single node that I'd like to run in parallel. Some of the rows will take much longer to compute than the most, so I'd like to use streaming for at least that node alone. Unfortunately, the node didn't run in parallel.

Christian Dietz told me that's because the token count is still 1. Do you already have plans to change that?




Hi Lorenz,

what about using the 'Parallel Chunk Start' and '-End' nodes from the KNIME Virtual Nodes extension?

Hi Martin,

when I have only one row inside the chunk that takes 5x as long as the others, the whole iteration has to wait for that. This is pretty much my use case, so I'd like to use streaming instead.

Any update on this?