Datawarehouse example

Can any one help me with an example of creating a data-warehouse and reporting using KNIME.
thanks in advance

1 Like

KNIME is not a data warehouse. You need to use a database to build one.

I meant an ETL process

For start

1 Like

Hi @KaiToRed!

This is a generic question but I will try to answer it somehow. Creating ETLs with Knime is possible and it is not to hard but it is not easy as well. Besides what @izaychik63 already provided I can share my experience and point out group of nodes which I used for ETLs.

As already said it is not too hard but not knowing some things will make your life a bit painfull :slight_smile: So Database node group has all the necessary nodes for reading from DBs and for writing to DBs. Also there are nodes which are preforming DBs operations directly on DB. What you have to decide is whether you will read all data from DB into Knime, preform data manipulation in Knime and then write it back to DB or you want to preform all operations and data manipulation directly on DB (of course this option is possible only in latter stages of DWH meaning for first step where you have to read data from one DB and write it to another). Performing data manipulation with Knime takes longer because you have to read and write data from DB and to DB which depends on your jdbc driver. Performing data manipulations directly on DB is much faster but you lose a bit of business logic and intermediate result. And you have to write SQL which is not a bad thing if you know how to of course. There is currently a new Database Integration node group being test (which you can try as well) so maybe you want to take a look on that as well.

Other used group of nodes are Workflow Control and Manipulation. There is much more things to tell but I will stop here and if any concrete questions of subject area around Data Warehouse and ETLs feel free to ask.



Thanks for your explanation.
I’ll try to do some manipulation.

1 Like