Reuse SQL data downstream in workflow

I keep running into a scenario that I’m wondering if there’s a better way to handle.

At the beginning of a workflow, I grab some data out of MySQL, I use that data to do some stuff, and then later in the workflow I need to do a couple comparisons to that original dataset, mostly as a QA kinda thing.

So, I can connect the first MySQL node, but then I have this long line running across the entire workflow. This is totally just a cosmetic issue, I understand that. Since I often work with small datasets, I’ll sometimes just copy the MySQL node and pop it in down stream again, but of course if I have to update the MySQL I have to remember to do it in both places.

So, I was just curious if there is a better way of either reusing data from one node later in the worklow or visually organizing the workflow.

I have some very long linear workflows, so it’s possible I just need to learn better organization methods. I’ve been using Knime for a little over a year now, so I’m starting to move away from the quick and dirty and create better workflows - largely due to this awesome community!

1 Like

You can wrap you nodes coming after SQL Reader into metanode.

1 Like

You can connect the output port of your first MySQL node to more than one input port, so just connect it also to any additional node later in the workflow

1 Like

@karlfloh, that’s what I have been doing. The little ocd in me was wondering if I could do a better layout to avoid the connection lines running through all the nodes in the middle. I’ve recently started using the annotations to document and organize the workflows so now the lines running through the middle of the workflow stand out more.

Totally a cosmetic thing at this point :slight_smile:

@izaychik63, funny you mention that. I just had some info on metanodes come up in some of my reading. I’m a little anxious about hiding all the work nodes right now, but I was starting to see how that would be useful here. I’m still at a point where I like seeing all the individual nodes, and I’m still making adjustments here and there, but I think this is what I will be using soon.


If you klick on the connection line you get handles on the line. If you then position the mouse pointer on such a handle, the mouse pointer changes to a cross with 4 arrows - hold down the left mouse button an drag the line away from crossing other nodes.

@izaychik63 WHAT??? How did I not know that was a thing. That’s exactly what I was looking for, thanks!


Just mark some sequential nodes following DB Reader right click and create metanode.

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.