Azure Analysis Services Overview
This write up assumes the reader is familiar with general Azure concepts, Analysis Services and OLAP technology, and development environments.
Please refer to my presentation for some introductory information as a supplement to this post.
This post will cover the benefit of using such a product within an organization, migration practices if moving from on-premises, considerations when comparing to on-premises implementations, automation tools to optimize an organization’s Azure Analysis Services build, and considerations for return on investment.
Business intelligence structures are ultimately striving for transparency, efficiency, and usefulness of meaningful data to run a business. Azure Analysis Services allows for this through informative and fast reporting underpinnings.
For example, here is a Power BI map that is plugged into an Azure Analysis Services instance:
It can provide insight as to where the company is selling the most product and possibly why it is not selling to cities that report lower profits, all with the backend power and speed of Azure’s massive resource pool.
A standard, albeit very simplified, architecture of data flow in an organization typically looks like this:
In this model, data sources are flowing into a pool or data lake in its raw format, after which data gets transformed into “business objects”, or tables that represent business entities properly. Those entities are staged for models in the Azure Analysis Services engine, before being read by reporting mechanisms.
This is much more easily said than done. For the purposes of this post, it is assumed that this is all working properly for use of the sample data set in the Azure Analysis Services layer.
A company using this model can retrieve insights and trends, save some space and time, have reports on portable devices, and explore data outside of the organization relatively easily.
Azure Analysis Services works with a variety of reporting tools, each with their own flavor of presentation, such as SharePoint, Power BI, and other third-party applications.
This is a strong case to have Analysis Services within a business intelligence architecture, in general.