Frequently asked questions about Azure Cosmos DB for MongoDB in vCore architecture

  • FAQ

APPLIES TO: MongoDB vCore

What options does Azure Cosmos DB offer for MongoDB users?

Azure Cosmos DB is a fully managed NoSQL and relational database for modern app development.

Both, the Request Unit (RU) and vCore-based Azure Cosmos DB for MongoDB offering make it easy to use Azure Cosmos DB as if it were a MongoDB database. Both options work without the overhead of complex management and scaling approaches. You can use your existing MongoDB skills and continue to use your favorite MongoDB drivers, SDKs, and tools by pointing your application to the connection string for your account using the API for MongoDB. Additionally, both are cloud-native offerings that can be integrated seamlessly with other Azure services to build enterprise-grade modern applications.

How do I choose between Azure Cosmos DB for MongoDB (RU) and Azure Cosmos DB for MongoDB (vCore)?

Here are a few key factors to help you decide between the two options.

  • Choose RU-based if:
    • You're building new cloud-native MongoDB apps or refactoring existing apps for cloud-native benefits.
    • Your workload has more point reads (fetching a single item by its _id and shard key value) and few long-running queries and complex aggregation pipeline operations.
    • You want limitless horizontal scalability, instantaneous scale up, and granular throughput control.
    • You're running mission-critical applications requiring industry-leading 99.999% availability.
  • Choose vCore-based if:
    • You're migrating (lift & shift) an existing MongoDB workload or building a new MongoDB application.
    • Your workload has more long-running queries, complex aggregation pipelines, distributed transactions, joins, etc.
    • You prefer high-capacity vertical and horizontal scaling with familiar vCore-based cluster tiers such as M30, M40, M50, and more.
    • You're running applications requiring 99.995% availability.
    • You need native support for storing and searching vector embeddings.

Are there billing differences between the RU and vCore architectures in Azure Cosmos DB for MongoDB?

The RU and vCore services have different architectures with important billing differences.

The RU service variant is a multitenant service that instantly allocates resources to your workload, aligning with storage and throughput requirements. Throughput for this service is based on request units. This service is optimized for users who prefer to pay fixed (standard provisioned throughput) or variable (autoscale) fees corresponding to Request Units (RUs) and consumed storage. Actual RU charges depend on the selected model: provisioned throughput (standard or autoscale) or serverless.

The vCore service variant creates dedicated instances that utilize preset CPU, memory, and storage resources, which can dynamically scale to suit your needs. This service is optimized for users who prefer to pay a consistent flat fee based on compute (CPU, memory, and the number of nodes) and storage.

Related content

Next step

Frequently asked questions about Azure Cosmos DB