We are pleased to announce the general availability of Spring Cloud Data Flow 1.4. You can download the Local, Cloud Foundry and Kubernetes releases from the Spring repository.
Here are the release highlights.
Spring Cloud Data Flow Ecosystem
Spring Cloud Stream 2.0 is nearing the GA release milestone (scheduled: March 29th). There have been a lot of stability and documentation updates. Given the active community testing across all the different parts of the framework, we were able to work on a few last minute usability improvements, too.
Likewise, Spring Cloud Task is nearing its 2.0 GA release as well (scheduled: April 11th). The team is focused on stabilizing the codebase in preparation for the upcoming major release.
A minor release of Spring Cloud Skipper with improvements to Kubernetes and Local deployment via the Kubernetes and Local Deployers respectively was recently released. If you’re curious how to continuously deliver streaming pipelines in SCDF with Skipper, you can refer to the reference implementation presented at DevNexus recently (slides).
Stay tuned for Tensorflow Objection Detection as an out-of-the-box processor application. The PR (spring-cloud-stream-app-starters/tensorflow#9) was recently merged – see docs here. We are getting ready to publish a solution architecture involving this application.
Having fun integrating latest object detection @TensorFlow models with @springcloud Data Flow for real-time image processing. @martinkl, the @intensivedata is recognized as a book although the picture 😉 Also, i don't have microwave 😉 pic.twitter.com/k9IJTOlXNX
— Christian Tzolov (@christzolov) February 15, 2018
Thanks for all the feedback on the other out-of-the-box applications. We are active in the community forums; please keep them coming and consider contributing to the starter applications.
UI/UX Refresh
A big focus of 1.4 release is to refine further and improve the user-experience of the Dashboard.
-
App Registration: The bulk-registration workflow has been redesigned. A more lean and lightweight look and feel has been embraced in the Dashboard overall.
-
App Versioning: This release adds the ability to register multiple versions of the same application from within the Dashboard.
-
Defaulting App Versions: Switching “default” version of the application in the registry is just 1-click away.
-
More Space: The Dashboard users would appreciate the extra real-estate; especially, when designing the streaming/batch pipelines in the visual designer.
It is better to see it in action as opposed to describing in words. Here’s the quick-start guide to Local, Cloud Foundry or Kubernetes.
Stream Deployment Builder
Extending upon the UI/UX improvements, we have developed a brand new stream deployment builder. To deploy a stream, you don’t have to remember the application, deployment or global properties and the boring freeform-text-area anymore. The “builder” comes pre-filled with pull-down menu options, so you can quickly make a selection, apply the overrides either individually at each application level or as a global setting. There’s mutual exclusive toggling in place to help you with the property overrides.
Security
Based on the recent feedback, this release adds support for mapping the LDAP Active Directory Groups with the Roles in Spring Cloud Data Flow. The configurable overrides can be applied as Spring Boot properties. The translation, filtering, and mapping of SCDF Roles with the Active Directory definitions are automatically handled by the system.
Getting-started Improvements
Spring Cloud Data Flow is a lightweight Spring Boot application. To automate the provisioning of the server along with its associated companion servers such as Skipper and Metrics Collector onto Cloud Foundry or Kubernetes, we have a Pivotal Cloud Foundry Tile and a Helm Chart respectively. Both of the provisioning experiences will upgrade to the 1.4 releases shortly – stay tuned!
However, to bring similar automation to the Local Server, we have added a Docker Compose template in this release. A `docker-compose up` command would automatically provision the latest GA release of the Local-server, Kafka, and an embedded H2 database to quickly get started in the development environment. All of these options are customizable – change from Kafka to Rabbit or switch to a different RDBMS – it is just a few steps away.
A similar Docker Compose experience is now available to Spring Cloud Stream samples, too. It is a hassle to find a compatible messaging middleware and the installation of it merely to get started with the samples. We hope this automation simplifies that—do let us know what you think.
Join the Community!
We are entirely a distributed team spanning across eight time zones, so one of us will be always online. Reach out to us in Gitter, StackOverflow, or in GitHub. Lastly, please try it out, ask questions, and give us feedback. We welcome contributions!
