KNIME performance and the length of the workflow

Dear all,

I have encountered the following problem. As soon as a workflow becomes very large (about 300-400 nodes, counting also nodes in meta-nodes) it takes forerver for KNIME to load a new file and run it through the workflow. I would appreciate any ideas which may help to solve this issue! Thanks!

/Aleks

Hi,

I don’t know too much about tha internals but from a users perspective does this correlate with you memory usage? Is there less memory availble to that file reader then the first one?

Best regards,

Jay

Another thing just came to mind when I re-read your post. Are you rseeting the original file input node. If so is your delay while the workflow resets?

Thanks Jay, for asking all the right questions.
Can I add one?: what does “forever” mean? Are we talking many seconds, minutes, hours?
I could see a rerun of a workflow take a few seconds if you have hundreds of nodes
doing something with your data. Plus, what kind kind of operations do you perform,
simple preprocessing or more complex analysis tasks?

Thanks, Michael