Sort by:  

Authors: Jianbo Sun, Da Yin

tldr - powered by Generative AI

KubeVela is a modern software platform that simplifies delivering and operating applications across hybrid multi-cloud environments. The platform provides an infrastructure-agnostic layer to render or deploy software and fully observability for the whole platform. The main challenges of KubeVela are balancing extension and user experience, powerful orchestration, resource management, and large scale and stability.
  • KubeVela is an application-centric control plan that connects with traditional CI systems or modern Github and provides best practices for platform engineering.
  • The Open Application Model provides a consistent model for application delivery and infrastructure-agnostic model that allows developers to focus on components trades and deploy to different runtime platforms.
  • KubeVela balances extensibility and user experience by leveraging the Kubernetes API and CRD ecosystem, providing API systems, SDKs, and a UI console.
  • KubeVela provides powerful orchestration by allowing data passing for application components dependency, overriding parameters for multi-environments delivery, defining rollout and traffic splitting ways, and checking metrics with quality gates to ensure safe delivery.
  • KubeVela manages resource management by sharing resources in multiple applications, defining the replicas of application at initialization while working with an HPA later, and defining different garbage collection policies for data storage and workload metadata.
  • KubeVela ensures large scale and stability by horizontally scaling the controller, improving performance for different scenarios, and providing a catalog of add-ons that can connect to the whole community.
Authors: Daniel Higuero

tldr - powered by Generative AI

The presentation discusses the use of Kubernetes and Helm in a multi-cluster approach, as well as the integration of GitHub with Flux CD add-on. The speaker also touches on the importance of flexibility in implementing the OEM specification.
  • Teams proficient with Kubernetes and Helm work well together, but onboarding new members can be challenging
  • Multi-cluster approach involves creating a customized entity with standard CRDs and deploying Helm charts as components
  • Flux CD add-on enables integration with GitHub and deployment of Helm charts as components
  • Flexibility is important in implementing the OEM specification, as it defines the minimum requirements but allows for additional features to be added by the runtime