Food For Tech | Meta Driven Tabular model
Author: Benito van Breugel
Data is everywhere and always changing, since everything is data, it is important to have an information platform that can cope with changes in a scalable, flexible, and robust manner. One of the solutions available that serves the business with insights is called an Analysis Services Cube model (OLAP capability). It can run both on-premise and in the cloud. Therefore, this service delivers great business value, drives better analyses and decision making without impacting your ICT strategy.
Having an OLAP capability will drive better analyses and decision making. Building an Analysis Services Cube model can be very complex and time consuming. Adding changes, small or large, can take quite some time when you manually need to develop, deploy, and process your changes across all your environments. Did you know that you can prevent this by generating, deploying and processing an Analysis Services Cube Model in an automated fashion?
At Food For Analytics we believe in a meta driven based approach across the board. Allowing us to adapt to changes quickly in our Scalable, Flexible, and Robust information platform.
Our Cube automation framework is based on C# and works with the Tabular Object model. Based on this, as part of our information-platform-offering, we have built a solution that automatically will generate, deploy, and process all artifacts that comprise an Analysis Services Cube Model.
Putting it to the test, we have enabled this on the AdventureworkDW2017 database. The database can be downloaded here. Based on the database-objects in the database, we have selected the object “dbo.FactFinance” as a starting point to automatically generate our Analysis Services Cube Model, including related dimensions.
The key to generate the model automatically is by populating the meta-base with the correct meta-data. The meta-base is the foundation of our automation framework. Based on the object selection (e.g. “dbo.FactFinance”) we can store all related tables and columns in one go, including their relationships. The only manual addition in the meta-base is the addition of measure(s) and hierarchies as they vary from organization to organization. In our example, we defined a measure on the column "Amount” in database-object “FactFinance”: (SUM(FACTFINANCE[AMOUNT]), which is a simple DAX equation. Once added, we initiate the Food For Analytics automation framework to generate, deploy, and process our Analysis Services Cube Model as defined in the meta-base.
Once we start the automation framework, we will get an output window to show the progress. First, if present, we remove all dependencies on the existing Analysis Services Cube Model.
Next we generate and deploy the Analysis Services Cube Model steps by step. At the end we will process the full Analysis Services Cube Model, making it available right away for the end-users.
Once done, the (new / updated) Analysis Service Cube Model is directly available and ready to be used by business in the defined endpoint (on-premise or cloud). As depicted in the figure above, the whole process has been executed in only 31 seconds for 2 Analysis Services Cube Models! (runtime may vary based on data volume). The generated Analysis Service Cube Model “ADVENTUREWORKS” (including the Finance domain) is shown right.
With the Food For Analytics automation framework we automate repetitive processes such as development, deployment and processing through meta-data. This allows solutioning teams to focus on what really matters; the business value the information platform delivers.
By automating repetitive processes, we reduce the manual involvement and potential human error to a bare minimum. Allowing your team to drastically shorten it's time to market to deliver actionable insights for your organization.
Adapting to change or adding new features to your Analysis Service Cube Model only requires you to add or change the meta-data in the meta-base and initiate the framework to regenerate, deploy, and process a new version of your Analysis Service Cube Model again.
When needed or requested, we will use this method to quickly unlock valuable insight for your business. Together we will guide you and help your drive better decision making with the use of our Scalable, Flexible and Robust information platform.
Interested? Feel free to reach out in case you want to experience the full demo with your own data!