Azure Cosmos DB for PostgreSQL vs Azure Database for PostgreSQL

Question

We are evaluating Azure Cosmos DB for PostgreSQL vs Azure Database for PostgreSQL. Is there a performance comparison between the 2 databases?

Answer 1

Hi, @devla.vsingh Azure Cosmos DB for PostgreSQL is situated for new development of cloud-native applications that have a need for elasticity to run at any scale. It’s highly scalable but can be used to run applications at any scale because of its elasticity. If you are building a new solution and want to be able to grow down the road, this would be a great place to start. Most notably Cosmos DB for PostgreSQL is a distributed database. Meaning queries can be routed to specific workers producing results in a map-reduce pattern where the work is done in parallel (greatly speeding up execution) and the results are then aggregated transparently for the requestor. This also allows for great IO scalability for reads but most notably for writes (especially if we factor in the ability to query/write to any node in the cluster). The product feature set puts Cosmos DB for PostgreSQL as a great candidate for multi-tenant SaaS applications and IoT-like time series data aggregation (due to writing scaling). In both scenarios, we allow unbounded horizontal scaling by adding more nodes and vertical scaling by increasing the storage, CPU, and memory available for each node in the cluster. Getting the most out of Cosmos DB for PostgreSQL requires the application developer to distribute the schema by selecting a distribution column for tables, one also should make sure that the application is properly using the distribution columns in queries.