Today we announced an update to Kalido Information Engine version 9. One of the awesome new features is its ability to exploit existing models, taxonomies and other related assets like business glossaries to accelerate delivering on a data mart migration and consolidation. We’re seeing more and more companies begin to tackle this problem – or continue to struggle with it. As I’ve posted before, traditional approaches to building data warehouses don’t cut it anymore because they aren’t inherently agile, and that’s what’s needed to deliver fast AND be able to accommodate future needs.
So what exactly is new? In Kalido Information Engine version 9 SP1, we can now import existing physical and logical models, data dictionaries and taxonomies into the Kalido Business Information Modeler. These can be in CSV format or any OMG CWM 1.x format. What comes out is your source model in Kalido’s business model format. This includes all the names and labels from your source model – so it likely won’t be easily understood by your business user counterparts. So, this release also automatically does name and label management, taking your abbreviation standards file and converting it from techno-speak into normal language. This means your naming standards can be automatically applied.
Once you have imported your first data mart model, you can then use the capabilities in the Kalido modeler to refactor the model to rationalize the model or accommodate the business’s newest requirements. The graphical nature of working with the Kalido modeler makes it easy to refactor the model and put it into a state that is easy to alter for handling future requirements.
This is really easy to do for your first data mart. But what if you bring in a second one? In a traditional approach, folding in a second model to an existing one could break many things in the first mart. As you refactor in the second model, at the core there may be keys that need reassignment, new tables , associations and the like. For example, in data mart 1, you only track direct sales. But in data mart 2, you track indirect sales. You might therefore need to route a new sales fact to a new element in a customer dimension. With Kalido, this is no problem because the powerful modeling technique allows the model to grow incrementally and the model-driven automation handles all the impact analysis, physical schema refactoring as well as data refactoring for you. You’ve just saved a bunch of work and resources on NOT sniffing out loads of ETL jobs to reconfigure.
As a result you’re now off to a much faster start to migrating and consolidating your data marts to a new, more agile warehouse foundation that can deliver more consistent information for use in business analytics.
Another key new capability in SP1 is support for Oracle Exadata. We’re really excited about this because we have seen significant performance gains running Kalido on Exadata. In addition to the performance, the Oracle Exadata Database Machine is also cool because it handles mixed workloads and can therefore handle both OLTP as well as analytical processing. For customers looking for a single machine that can handle data mart consolidations, the combined solution of Kalido on Oracle Exadata is a compelling option.