GPM, Azure Cosmos DB, Microsoft
Globally Distributed Databases Made Simple
Public clouds are quickly making massive-scale computing capabilities available to an ever-larger population of developers and data professionals. These computing capabilities are no longer a playground restricted to a small handful of large-scale internet services organizations. In 2014, I gave my first talk at PASS summit on Cloud Databases 101 – a lot has changed since then. Globally distributed databases in the cloud have created a dramatic impact in today’s business landscape. In my talk this year, I will provide a deep dive understanding into Azure Cosmos DB, Microsoft’s globally distributed, multi-model database service that was 7 years in the making. We’ll take an under the covers look at the system and discuss various design trade-offs that had to be made to allow customers to elastically and horizontally scale both throughput and storage globally. We’ll also discuss how the system can provide single digit millisecond latencies with 99.99% high availability. I will also share some of our real-world experiences in operating a globally distributed database service and will touch on how Azure Cosmos DB compares to the competition (including Google Spanner, AWS DynamoDB, Cassandra, MongoDB and others). How Cosmos DB relates and works with other technologies like stream analytics, Azure Machine Learning, predictive analytics will also be covered. You’ll walk away with a simple yet comprehensive understanding of Cosmos DB and an appreciation for the beauty of the system.
Rimma has previously worked at Microsoft Gray Systems Labs and Microsoft Research and holds a PhD in Computer Science from Purdue University and an MBA from the University of Chicago.