D365 FinOps & Power BI
Our client in the travel industry presented us with the challenge of extracting data from Dynamics 365 Fin Ops for visualisation in Microsoft Power BI. This resulted in us having to write custom objects to expose the data from Dynamics 365. With potentially billions of rows of data, following recommended architectural patterns is extremely important to ensure optimal effeciency and minimal impact to production systems.
Here are the challenges we’ve faced to date and what we’ve learned so far.
· Accessing data: As an ERP system that can cater for large complex organisations, D365 is built to process and store vast amounts of data. As such, the underlying data structures are dynamic, customisable and stored in a highly normalised form, making it too complex to read directly, not to mention any performance impacts of reading from an online transactional database.
· Documentation & new developments: The recommended options for data extraction are documented but are in a state of constant change. Some current methods for accessing data will be phased out while new features are not available for production use yet.
· Custom development can be required: Some options involve a level of D365 development that is not part of the typical business intelligence developer's skill-set (X++, D365 objects and knowledge of the D365 DevOps build process, etc).
D365 vs Entity Store vs Data Entities vs OData vs Data Lake...
We found it very easy to get lost in the ocean of potentially outdated documentation, online articles, and sometimes conflitcting approaches to working with D365.
Entity Store (AxDw)
Entity Store (Data Lake)
Data Entities, OData & BYOD
Data Entities to Data Lake (preview)