Nodes with KubeEdge and Case Study Yulin Sun, yulin.sun@huawei.com; Li Xing, Li.xing1@huawei.com; Seattle Cloud Lab, Huawei R&D USA, Bellevue WA Agenda • Edge scenarios/characters • KubeEdge Architecture • Heterogeneous hardware • Memory: 64MB ~ 100+ GB • CPU: AMD64/ARM/MIPS… Manage Edge Nodes with KubeEdge • Goal • Manage Edge Nodes together with Nodes in cloud as one Cluster • Address the Edge special Node KubeBus App Engine EdgeMetadataService ... EdgeController Cloud Edge ETCD K8S Node2 KubeEdge VPN KubeBus • Scenario • Edge Nodes have no external IP and running behind NAT • One VM in Cloud

KubeEdge Demo Session Jun Du Huawei Cloud Cloud To Edge Low Latency • AR/VR: ms level • Industrial Control: us level High Availability • Availability: > 99.999% • Response Time: Predictable • Result: Evaluate and extend existing CLI tools to manage K8S clusters running in remote edge locations. KubeEdge A Kubernetes Native Edge Computing Management Framework. Mission: Let Cloud manages edge nodes Abundant Applications/Protocols Support • Simplify Device Access KubeEdge Architecture Overview KubeEdge(Edge Part) OS + Runtime KubeEdge(Cloud Part) App Devices Driver/SDK Volume Configmap Pod Prober

深度解析CNCF社区⾸首个基于Kubernetes的边缘计算平台KubeEdge� 向新勇� https://github.com/edisonxiang� Introduce� ➔ 华为开源社区⼯工程师� ➔ KubeEdge社区Member� ➔ Kubernetes社区Member� ➔ OpenSDS社区Memeber� ➔ OpenStack社区数据保护项⽬目联合发起⼈人� OpenStack社区数据保护项⽬目联合发起⼈人� Outline� ➔ 边缘计算 & 应⽤用场景 & ⾯面临的挑战� ➔ Why KubeEdge & 基础架构 & 设备管理理 & 实战� ➔ 后续规划 & 社区贡献 & 技术交流� 边缘计算� 云计算是集中化的，离终端设备（如摄像头、传感器器等）和⽤用户较远，对于实时性要求⾼高的计算需求，把计算放在云上会引起较⻓长的⽹网络延 时、⽹网络拥塞、服务质量量下降等问题 。� ➔ 管理理：边缘节点的资源管理理与边缘应⽤用⽣生命周期管理理。� ➔ 扩展：⾼高度分布和⼤大规模的可扩展性。� ➔ 异构：边缘侧异构AI硬件接⼊入。� Why KubeEdge� KubeEdge通过更更优的架构和技术实现，完美应对边缘计算遇到的挑战：� ➔ 通过将AI能⼒力力、⼤大数据能⼒力力等延伸到边缘，解决与云上服务的数据协同、任务协同、管理理协同、安全协同诉求。�

com Advancing the Tactical Edge with K3s and SUSE RGS Ki Lee Vice President Booz Allen Hamilton KubeEdge and K3s seemed the most natural starting point, given the device-centric use case. After assessing many stayed focused on the Cloud and had no clear strategy for edge deployments. The team liked KubeEdge, but with the profile of its customers front-of-mind Booz Allen wanted to work with a company

Security policy configurations across multiple clusters Edge computing Deep integration with KubeEdge; K3s supported; No feature available and OpenShift Sponsoring K3s project as the solution