Dapr v1.0.0-rc.2 is now available
We are excited to announce v1.0.0-rc.2, the second release candidate for the official production ready v1.0 of Dapr. Last month we released the first release candidate for Dapr v1.0 and it is great to have a new release candidate with additional improvements, updates and fixes in just a few weeks.
The improvements in this release came out of issues opened by the community which has helped test the previous release candidate as well as contributed code changes to address these issues. Once again, we encourage the Dapr community to upgrade to the second release candidate and identify problems and open issues.
We expect a third release candidate early next year.
Installing and using v1.0.0-rc.2
As with the previous release candidate, an explicit opt-in to download and install is required. The instructions to upgrade to v1.0.0-rc.2 can be found in the release notes, as well as in the v1.0.0-rc.2 docs.
If you are using the Java or .NET SDKs, the new releases include some breaking changes see these links for more information:
These are some highlights for this release:
- Added the ability to share state between different applications using prefix strategies. Visit this link to learn more.
- Added a bulk secrets API to get all the secrets at once from a secret store that the app has access to. Implemented in the secret components.
- Actors in Java SDK communicate to Dapr sidecar via GRPC - This allows better performance for invocation of Actor methods, timers and reminders. ActorProxyBuilder is now closeable, holding an instance of gRPC’s ManagedChannel class to be closed. See upgrade instructions for the Java SDK for more information.
- Deprecation of exporters component (issue #2337) - Tracing is now part of Dapr runtime, and tracing settings have moved into configuration. See this doc for information about the new tracing settings. For more detailed instructions on setting up tracing in new Configuration, check out the observability quickstart sample as well as guides for Zipkin, Jaeger, New Relic and OpenTelemetry trace backends such as Azure Application Insights.
- Service invocation performance test results now published and available at the Dapr performance website.
- Full review on the Dapr CLI to make command parameters consistent, update descriptions, provide more examples and adding capabilities such as getting the output for configuration.
- Improvements to the Dapr metrics dashboards and Dapr operational dashboard.
- Round robin load balancing of mDNS requests - This supports multiple app instances for a single standalone app id with round robin load balancing across them for environments that support mDNS, such a physical machines.
- Added the option to specify whether Dapr should ignore errors when loading a component and continue initialization. See this link for the new ignoreErrors field.
- Dependency injection (DI) for .NET actors.
- Added Java and Python versions for the Twitter sentiment processor sample. See details here.
See a detailed description of changes, fixes and new features in the Dapr v1.0.0-rc.2 release notes. As always, it is recommended to pay careful attention to the breaking changes section before upgrading.
We hope you try out this release candidate and raise issues on GitHub as well as reach out to us to ask questions on Discord or via Twitter - just give us your thoughts, we’d love to hear from you!