Question

A retail company stores order documents in an Azure Cosmos DB for NoSQL container. The architecture team wants every new or updated order to be streamed to multiple independent downstream systems: a real-time fraud-detection service, a Spark streaming job, and a long-term data lake archive. Each consumer must process the same stream independently at its own pace, and the solution should require minimal custom polling code against Cosmos DB. Which approach best meets these requirements?

Accepted Answer

Use an Azure Function with a Cosmos DB trigger (change feed) that publishes each changed document to an Azure Event Hub, then let each downstream system consume via its own Event Hubs consumer group.

Answer

Have each downstream system directly query the Cosmos DB container on a schedule using the API for NoSQL and filter by a lastModified timestamp.

Answer

Enable analytical store with Azure Synapse Link and point all three downstream systems at the analytical store.

Answer

Configure a single Azure Function change feed processor that writes directly to the data lake, and have the other two systems read the archived files.