Designing analytical stores
Analytical stores could either be SQL or NoSQL data stores deployed in the data lake Serving Zone. The main job of an analytical data store is to serve the data generated by the data transformation pipelines to the BI tools in a fast and simple manner. Analytical stores are usually subjected to ad hoc querying from business analysts and other end users. As such, these stores need to perform really well with random reads. Azure provides a variety of storage technologies that can cater to these requirements. Here are some of the most important ones:
- Azure Synapse Analytics (Serverless/dedicated SQL pools and Spark pools)—Synapse Analytics provides both SQL pools and Spark pools. Among these, the SQL dedicated pools are massively parallel processing (MPP) data warehouses and are usually ideal for a majority of analytical store situations. Serverless SQL pools can be used for ad hoc querying. Spark pools, on the other hand, can support analytical...
