In Ajilius 1, we used Python as a scripting language. Ajilius 2 switched to Java and Groovy. We generated scripts that mixed the scripting language and SQL to emulate stored procedures.

We got pushback on this approach from a number of prospects, probably because we didn’t always do a good job of explaining the relationship between the scripting language and the SQL. Prospects saw the Groovy syntax, for example, and thought that they would have to learn a new language to build their data warehouse.

Looking back, we can see why.

Here is an Ajilius 2 script:

Here is the Ajilius 3 script for the same table:

As you can see, the SQL script in our new version is immediately familiar to a data warehouse developer.

This isn’t to say that we’re abandoning scripting, because Ajilius 3 introduces this feature as SQL + Scripts.

You’ll still have the option to export and package the SQL for execution as Groovy scripts, Stored Procedures (where supported), PowerShell scripts and more. The difference to earlier versions is that this is an option, not the standard.

An added bonus from this approach is that scripts run through the Ajilius scheduler no longer have to go through the overhead of instantiating the scripting engine and loading/compiling the script before execution. As a result, you get much faster execution of your transformation processes.

Ajilius. Listening to customer feedback to make us better.


Leave a Reply