So far, each time you got an error, you had the opportunity to discover what kind of error it was and fix it. This is quite different from real scenarios, mainly for two reasons:
- Real data has errors—a fact that cannot be avoided. If you fail to heed it, the transformations that run with test or sample data will probably crash when running with real data.
- In most cases, your final work is run by an automated process and not by a user from Spoon. Therefore, if a Transformation crashes, there will be nobody who notices and reacts to that situation.
In this section, you will learn the simplest way to trap errors that may occur, avoiding unexpected crashes. This is the first step in the creation of transformations ready to be run in a production environment.