Question

Your team is starting a brand-new analytics initiative on top of an existing Azure Cosmos DB for NoSQL account. The goal is to run large-scale analytical queries and build Power BI reports over operational data without impacting transactional RU consumption, and without building or maintaining ETL pipelines. Following current Microsoft recommendations for new analytics projects, which approach should you choose?

Accepted Answer

Mirror the Cosmos DB database into Microsoft Fabric, then query the mirrored data through the Fabric OneLake SQL analytics endpoint

Answer

Enable Azure Synapse Link on the containers and query the analytical store from a Synapse serverless SQL pool

Answer

Provision a dedicated SQL pool in Synapse and run scheduled Azure Data Factory copy jobs from the transactional store

Answer

Read the change feed with an Azure Function and write each document into an Azure SQL Database for reporting