A List of the Cosmos DB APIs
As well as scalability, availability, and low latency along with global distribution mentioned in the previous lesson, Cosmos DB had a design goal of supporting multiple data models.
In your IoT technology company, you have customer-facing databases hosted on several different systems including MongoDB, Cassandra, Microsoft SQL Server, and Gremlin. This is because you grew by acquiring other IoT companies with complementary technologies, but each original company had chosen a different backend technology to host its telemetry and other databases. If you're going to migrate to the cloud, you need a solution that will support all these databases without requiring you to rewrite software for your devices, which would be prohibitively expensive.
Here, you'll learn about the support that Cosmos DB includes for disparate NoSQL data models.
How Cosmos DB supports NoSQL data models
To support multiple data models, Cosmos DB uses an API for each data model that provides projections of containers and items as familiar objects to that data model. This enables existing NoSQL databases to be migrated to Cosmos DB with minimal disruption while taking advantage of the benefits of the underlying Cosmos DB engine. For more information, see A technical overview of Cosmos DB
Cosmos DB APIs
Cosmos DB includes multiple APIs that provide mappings for native applications to the Cosmos DB engine. These include the following, but the list will expand as more APIs are developed, including the etcd API for Kubernetes.
SQL
The SQL API is the core API for Cosmos DB and the SQL API was originally the DocumentDB interface. If you are creating a new database with no need to support pre-existing apps, it makes most sense to use the SQL API.
This course addresses the needs of those who want to migrate Casandra and MongoDB solutions to Cosmos DB. You can choose to migrate your database to the SQL API of you prefer its features. Many of the migration tools, including the Azure Cosmos DB data migration tool and Azure Data Factory can help you with this migration. However, to migrate the whole solution you'd have to rewrite your client applications to use the SQL API instead of the Cassandra or MongoDB APIs. For this reason, this course concentrates on Cassandra and MongoDB APIs in Cosmos DB, and not SQL API.
Cassandra
The Cassandra API enables apps written for Apache Cassandra to interface with the Cosmos DB engine. The apps can use Cassandra terms for objects, the Cassandra Query Language (CQL), Cassandra drivers, and Cassandra tools. Little or no recoding is required for these apps after you migrate your database into Cosmos DB.
MongoDB
The Cosmos DB MongoDB is compatible with the MongoDB wire protocol version 3.2 with version 3.4 in preview. Applications created to support these protocols should be able to connect to the Cosmos DB MongoDB API. Little or no recoding is required for these apps after you migrate your database into Cosmos DB.
Gremlin
Gremlin is the Apache TinkerPop graph traversal language. The Gremlin API enables you to create graph entities in Cosmos DB.
Table
The Cosmos DB Table API supports applications created for Azure Table storage while providing the scalability and availability advantages of Cosmos DB.