Problem of Guided_Analytics_for_ML_Automation running on the KNIME Server

I have run the Guided_Analytics_for_ML_Automation from the HUB examples in my server successfully.
But with my own csv file, about 130 columns and 500k rows, when do training the model, can not finish the workflow.
Will show the “Java heap size” error in the log.
How to increase this size on the KNIME Server?
I have update the “-Xmx64g” in the file knime.ini under the “knime_executor” foder.

Hope someone can help me. Thanks.

Hi whzhang,

Just to confirm, what is the total amount of memory available on your KNIME Server machine? The -Xmx value should be a fair bit lower than the total memory.


We have 128G memory in our KNIME server.
Is there any special configuration of Node in the workflow?

Can you try to set a lower value for -Xmx? I suggest 48 GB. Just want to make sure whether heap issues are due to executor rotation or if there is something else going on.

sure, I will have a try.