Question

A retail company is designing a container in Azure Cosmos DB to store millions of customer orders. They want queries to perform well and want their data and request throughput spread evenly across physical partitions to avoid 'hot' partitions. Which design decision most directly helps achieve this even distribution?

Accepted Answer

Selecting a partition key with high cardinality, such as customer ID, that spreads requests across many values

Answer

Choosing the eventual consistency level instead of strong consistency

Answer

Enabling multi-region writes across several Azure regions

Answer

Using a partition key with a small number of possible values, such as order status