Pentaho Kettle insert Error Handling of step

0
votes

I am new to GeoKettle (Spoon) of Pentaho and I am currently rows from an Excel-file into my database. Now I want to avoid duplicates in my databasetable. That is why I want to insert only those rows into my database table which aren't there yet (to have only unique records in my database table).

And as far as I know, there are two ways to realize that. The first way I tried was with the Insert/Update step (I have disabled the Update functionality) and defined all the columns which have to be equal in order to insert the record or not. But it does not work. All records are still inserted into the database.

That is why I am trying to do the (according to pentaho) much faster option which is a "Table Output" with an "Update" error handling step as shown in the picture. enter image description here

As shown in the picture, the arrow which is pointing from "table output" to "update" is black. But I need a red dotted one for error handling of step . But I do not know how to create this. In tutorials I often see that there pops up a little window with 2 options like in the picture: enter image description here

But I do not get that popup. If I want to create a hop, I will have to mark both steps and do a right-click on one of them. So in which possible ways can I create such a red dotted arrow? In the end, it has to look like this:

enter image description here

Thank you so much in advance!!

1
mysqlpentahokettlepentaho-spoongeokettle

1 Answers

1
votes

You have a problem with your setup. Or with your version of the PDI. The functionality of an error step was introduced in V4 but fully implemented for all steps around V6.

Download a fresh PDI from SourceForge. V7.1 is really a robust and stable edition. Unzip and test.

By the way, what you want to achieve is know as the CRUD pattern. CRUD for Create, Read, Update, Delete. The step doing this the Merge Rows (diff) (in the Joins family). You tell the steps which columns to check, and it produce a new column with the value identical, changed, new, or deleted. You can them redirect the flow in a Switch / Case to do the appropriate action. Further information here (V4).

enter image description here