Azure Cosmos DB - Low Latency and High Availability at Planet Scale

Explore the internal architecture of Azure Cosmos DB in this conference talk from NDC London. Dive into the design of the storage engine, focusing on how high availability and scalability are achieved through partitioning and replication. Examine the evolution of the request routing gateway and its solutions to common multi-tenant cloud infrastructure challenges. Learn about performance as a feature and a culture, including measurement techniques and SLO considerations for maintaining low latency. Gain insights into the complex scalability challenges and tradeoffs involved in building planet-scale services like Azure Cosmos DB, which powers mission-critical services for Microsoft and Fortune 500 organizations. Discover how this fully-managed, multi-tenant, distributed database provides planet-scale capabilities and multi-model APIs for various database systems.

Azure Cosmos DB - Low Latency and High Availability at Planet Scale - Kevin Pilch