Hello everyone! Today I will bring you good news, Dapr (Distributed Application Runtime) version 1.15 has been officially released! For those who are not familiar with Dapr, Dapr is an open source, cross-platform runtime that simplifies many complex problems in the microservice architecture, such as service discovery, configuration management, state management, etc. So, what new features and improvements have been brought to Dapr 1.15? Let's take a look together!
1. Dapr Workflow (stable version)
First of all, Dapr Workflow officially entered the stable stage in version 1.15. This means you can now use Dapr Workflow with confidence to write long-running stateful applications without worrying about it "turning over". Workflow not only supports persistent execution, but also enables dynamic scaling, allowing your application to increase or decrease the number of replicas according to needs at runtime, while maintaining the persistence of tasks. This is a boon for applications that need to deal with a lot of complex business logic!
2. Dapr Actors Rewrite
In version 1.15, the runtime engine of Dapr Actors was also rewritten. Although the API level has not changed, this rewrite significantly improves the reliability and stability of Actor in large-scale and dynamic scaling environments. What's even better is that this rewrite also lays a solid foundation for the future implementation of more advanced Actor APIs and integration with other systems.
3. Scheduler Service (stable version)
Scheduler Service is a control plane service used to schedule jobs in Dapr, and it has also been officially stable in version 1.15. This means you can use it in production with confidence to manage your scheduled tasks or periodic jobs. Now, when you upgrade to version 1.15, all new Actor reminders will use Scheduler Service by default, instead of the previous Placement Service. Of course, you can also use configuration to choose whether to migrate existing Actor reminders.
4. Conversation API (Alpha version)
Dapr 1.15 also introduced a new Conversation API (currently in the Alpha phase). This API can help you interact with large language models (LLMs), such as OpenAI's GPT series models. Through the Conversation API, you can send prompts and receive responses to LLMs, and also support performance optimization and security features such as cache prompts to reduce calls and hide personal information (PII) for privacy.
5. SDK improvements
In terms of SDK, the SDKs of each language have been updated and improved to varying degrees. for example:
- .NET SDK: Added support for Jobs, Conversation and Streaming Pub/Sub APIs, while supporting .NET 9.
- Python SDK: Added support for streaming Pub/Sub API and Conversation API, and also introduced Mock actors for unit testing.
- Java SDK: Added Spring Boot integration, supporting automatic injection of DaprClient and DaprWorkflowClient.
- Go SDK: Added support for the Conversation API, and the writing and management of Workflow has also reached a stable state.
6. Upgrade Guide
If you are already using Dapr and want to upgrade to version 1.15, then don't worry, the Dapr team has prepared a detailed upgrade guide for you. Whether you are running Dapr on your local machine or deploying Dapr in a Kubernetes cluster, you can easily complete the upgrade by following the guide. Of course, there may be some incompatible changes during the upgrade process, so be sure to read the "Breaking Changes" section of the upgrade guide carefully.
Conclusion
Overall, Dapr 1.15 version brings many exciting new features and improvements. Whether it is the release of the stable version of Workflow or the mature application of Scheduler Service, we have seen the huge potential of Dapr in the microservice architecture. If you haven't tried Dapr, now is a good time! Come and experience this powerful microservice runtime!
Finally, let’s say a joke: After using Dapr, your code volume may decrease, but your hair volume will definitely increase. Version 1.15 is more recommended to drink with coffee. The official blog post has more details:/posts/2025/02/27/dapr-v1.15-is-now-available/。