The NoSQL database service Azure Cosmos DB can reduce the cost and latency of read workloads by integrating caching. Microsoft mentioned that in addition to speeding up and reducing costs, adding caching can simplify users’ architecture and application complexity. This integrated cache is built into the gateway dedicated to Azure Cosmos DB, and users need to make small adjustments to the application to use it, but the official mentions that this adjustment will not cause additional execution costs for the application. All the benefits of traditional caching.
Integrating the cache can reduce the database cost of many workloads. If the cache does not need to be updated, no matter the complexity of the query or the size of the data returned, it will cost 0 RU (Request Unit) to provide the read service. RU is the performance currency of Azure Cosmos DB, abstracting the system resources required to execute Azure Cosmos DB database operations.
Microsoft explained that this is more beneficial for relatively static data reading workloads, and when data is read more frequently than updated, there is a potential benefit to using caching. Executing the same data point read or query workload multiple times in a short period of time can also benefit from setting up caching, such as workloads where large files or complex queries cause high read costs, or when one or more A workload that has hot partitions due to repeated reads in a physical partition.
Azure Cosmos DB integrated cache features network isolation to ensure data security, users can configure dedicated gateways with virtual networks, IP filters and private connections, and users can monitor with complete and easy-to-read metrics Cache usage.