Cosmos DB: Everything you wanted to know about Microsoft Azure’s NoSQL offering in about 30 minutes

The ObjectSharp Podcast

In this, our very first episode, we talk with Shane Castle and Dave Judd all about Microsoft Azure's database service offering, Cosmos DB. We explore what Cosmos DB is, what problems it solves, how it stacks up to other databases such as SQL, Firebase, Mongo DB, etc., and why you may want to consider using Cosmos DB in the development of your next application. If you've heard about Cosmos DB but you're not really sure what's it all about, this is a great place to start.

Minutes:

• 0:00 - Introduction to the show
• 1:09 - A brief background on ObjectSharp and why we're doing a podcast
• 1:53 - Jeff and Nick talk a bit about who this podcast is for while introducing the topic of Cosmos DB
• 4:20 - Introduction to Dave Judd and Shane Castle
• 5:25 - What is Cosmos DB? Shane discusses Cosmos DB and its unique feature as an enterprise-class “multi-master” database as a service
• 6:50 - What are the core problems Cosmos DB attempts to solve? How is it different from other database platforms? Shane talks about how Cosmos DB solves issues of (1) time to market and (2) maintenance
• 7:55 - Consistency - Shane talks about Cosmos DB’s dynamically tunable consistency model
• 8:58 - Shane talks about Cosmos DB’s core APIs (SQL, Mongo DB, Cassandra, Gremlin (for graph databases), Table (for those using Azure table storage))
• 11:05 - Shane talks about Cosmos DB’s support for different languages (Java, Python, Go)
• 12:08 - Shane talks about Cosmos DB’s unique latency SLA
• 13:13 - Dave talks about Microsoft’s use of Cosmos DB within Microsoft and the benefits that come from that
• 13:46 - Dave talks about the implications of moving to Cosmos DB and database strategy, the importance of real-time data replication for app performance
• 16:05 - Dave talks about key differences between Cosmos DB vs other NoSQL, and between Cosmos DB vs SQL; how Cosmos DB query syntax is familiar to those coming from SQL (it’s SQL-like); and Azure’s “auto-tuning” of performance during inserts and auto-scaling features
• 22:30 - Shane talks about Cosmos DB in comparison to other NoSQL offerings, when you might want to consider Cosmos vs other NoSQL offerings
• 24:50 - Dave talks about Cosmos DB’s support for change feed and reactive programming; reactive pipes as native to Cosmos DB; Azure’s EventGrid service; how this allows for developers to more easily keep their data eventually consistent
• 28:16 - Dave talks about when is a good time for a company to think about Cosmos DB; valid use cases for it; business imperatives that might drive interest; the importance of the real-time replication, high availability, and having your writes close to the end-user
• 33:55 - Dave and Shane talk about Cosmos DB costs and pricing, how it differs from other options; Shane also talks about the importance of DR
• 37:58 - Getting started with Cosmos - how to get started and playing with Cosmos DB to see if it works for you
• 40:03 - Shane and Dave touch briefly on the related topic of serverless
• 44:32 - Closing and where you can learn more about Shane and Dave

This is our first episode. We'd love your feedback! Feel free to reach out to us on Twitter any time @teamobjectsharp.

Links to items mentioned in the show:

• Azure Cosmos DB
• Getting started with Cosmos DB for free
• ObjectSharp
• Dave Judd’s January 2019 Event on All Things Serverless
• Shane Castle’s February 2019 Event on Microservices, Containers & Orchestration (and how Docker and Kubernetes make it happen)