Icon for KubeEdge

KubeEdge

Open-source CNCF-graduated project that extends Kubernetes APIs to edge computing environments. Enables containerized application deployment and device management across cloud-edge infrastructure with ~70MB memory footprint.

Screenshot of KubeEdge website

KubeEdge is an open-source system that extends native Kubernetes container orchestration and device management capabilities to edge computing environments. Built upon Kubernetes, it provides core infrastructure support for networking, application deployment, and metadata synchronization between cloud and edge nodes.

The architecture consists of cloud-side components (CloudHub, EdgeController, DeviceController) and edge-side components (EdgeHub, Edged, EventBus, DeviceTwin, MetaManager, ServiceBus). This split design enables edge autonomy - edge nodes continue operating even when disconnected from the cloud, with metadata persisted locally using SQLite.

Key capabilities

KubeEdge enables developers to write regular HTTP or MQTT-based applications, containerize them, and deploy anywhere at the edge or in the cloud. The framework supports heterogeneous hardware architectures including x86, ARMv7, and ARMv8, making it suitable for diverse industrial IoT deployments.

The EventBus component provides native MQTT client capabilities for device communication, while DeviceTwin maintains device state synchronization between edge and cloud. Edge nodes require only ~70MB memory, enabling deployment on resource-constrained devices.

Limitations

  • Requires existing Kubernetes cluster as prerequisite; does not replace Kubernetes but extends it
  • Cloud-edge network instability can impact synchronization; designed for intermittent connectivity, not permanent offline operation
  • Device mapper ecosystem is smaller than dedicated IoT platforms like Eclipse Kura or proprietary solutions
  • Edge autonomy has limits: while metadata persists locally, some cloud-dependent features degrade during disconnection
  • Complex multi-node edge deployments require careful network planning for node-to-node communication
  • Limited built-in analytics capabilities; typically paired with separate stream processing or time-series databases for data analysis
Categories:

Share:

Kind
Platform
Vendor
CNCF
License
Open Source
Website
kubeedge.io
APIDeployment TypeLicensePlatformProtocol
Show all
Ad
Icon

 

  
 

Similar to KubeEdge

Icon

 

  
  
Icon

 

  
  
Icon