I’ve got some troubles with the analysis of my database of 34 million observations.
It takes many time to calculate a node (for example, a simple row filter node takes around one hour). How can I improve the performance of KNIME ?
What exactly is slow? The reading of the data from the database or the processing in KNIME? As far as I understand it’s the actual processing in KNIME?
What you can do is increase the RAM you give to knime in knime.ini. See this. You could try to up it to 5-6 GB. But for that you need to close all other things on your laptop including the browser.
Also depends what your row filter is actually filtering. A regex filter will probably be a bit more intense than a simple numeric range filter.
But regardless of that, 8 GB RAM simply is a little low if you are doing “data science stuff” with 34 million rows. On top of that your processor is from a 15W laptop, eg small form factor laptop
with a very low base frequency of 1.8ghz. You can’t expect wonders from this setup. And what you aren’t showing at all is your storage, If this laptop has a HDD (and not ssd) combined with the small amount of RAM the poor performance makes a lot of sense.