A portable, serverless, event-driven runtime that makes it easy for developers to build resilient, stateless and stateful microservices that run on the cloud and edge and embraces the diversity of languages and developer frameworks.
Dapr codifies the best practices for building microservice applications into open, independent building blocks that enable you to build portable applications with the language and framework of your choice. Each building block is independent and you can use one, some, or all of them in your application.
This is a Cloud Native Computing Foundation (CNCF) incubation project.
- Enable developers using any language or framework to write distributed applications
- Solve the hard problems developers face building microservice applications by providing best practice building blocks
- Be community-driven, open and vendor-neutral
- Gain new contributors
- Provide consistency and portability through open APIs
- Be platform agnostic across cloud and edge
- Embrace extensibility and provide pluggable components without vendor lock-in
- Enable IoT and edge scenarios by being highly performant and lightweight
- Be incrementally adoptable from existing code, with no runtime dependency
Dapr injects a side-car (container or process) to each compute unit. The side-car interacts with event triggers and communicates with the compute unit via standard HTTP or gRPC protocols. This enables Dapr to support all existing and future programming languages without requiring you to import frameworks or libraries.
Dapr offers built-in state management, reliable messaging (at least once delivery), triggers and bindings through standard HTTP verbs or gRPC interfaces. This allows you to write stateless, stateful and actor-like services following the same programming paradigm. You can freely choose the consistency model, threading model and message delivery patterns.
Dapr runs natively on Kubernetes, as a self-hosted binary on your machine, on an IoT device, or as a container that can be injected into any system, in the cloud or on-premises.
Dapr uses pluggable component state stores and message buses such as Redis and gRPC to offer a wide range of communication methods, including direct dapr-to-dapr using gRPC and async Pub-Sub with guaranteed delivery and at-least-once semantics.
Writing high-performance, scalable and reliable distributed application is hard. Dapr brings proven patterns and practices to you. It unifies event-driven and actors semantics into a simple, consistent programming model. It supports all programming languages without framework lock-in. You are not exposed to low-level primitives such as threading, concurrency control, partitioning and scaling. Instead, you can write your code by implementing a simple web server using familiar web frameworks of your choice.
Dapr is flexible in threading and state consistency models. You can leverage multi-threading if you choose to, and you can choose among different consistency models. This flexibility enables to implement advanced scenarios without artificial constraints. Dapr is unique because you can transition seamlessly between platforms and underlying implementations without rewriting your code.
- Event-driven Pub-Sub system with pluggable providers and at-least-once semantics
- Input and output bindings with pluggable providers
- State management with pluggable data stores
- Consistent service-to-service discovery and invocation
- Opt-in stateful models: Strong/Eventual consistency, First-write/Last-write wins
- Cross platform virtual actors
- Runs natively on Kubernetes using a dedicated Operator and CRDs
- Supports all programming languages via HTTP and gRPC
- Multi-Cloud, open components (bindings, pub-sub, state) from Azure, AWS, GCP
- Runs anywhere, as a process or containerized
- Secrets management to retrieve secrets from secure key vaults
- Rate limiting
- Built-in Observability support
- Lightweight (58MB binary, 4MB physical memory)
- Runs as a sidecar – removes the need for special SDKs or libraries
- Dedicated CLI – developer friendly experience with easy debugging