P2CODE: A programming platform for developing and deploying edge and IoT apps
summary
P2CODE delivers a Kubernetes operator scheduler built on Open Cluster Manager that accepts a custom P2CODE scheduling manifest, auto-bundles related manifests (configs, secrets, PVCs), and sets up scupper links when cross-cluster services must stay connected. Edge rollout uses “cubit” bootable containers to run K3s on NVMe/SD IoT nodes; two testbeds (OpenStack/OpenShift in Patras, migrated vSphere→AWS in Milan) feed a federated Prometheus for telemetry. Demoed use cases: exoskeleton operator monitoring and drone+UGV coordinated search-and-rescue.
AI-written summary · about
chapters (15)
- 0:00P2CODE platform intro
- 1:06Project background
- 2:35Key features
- 4:21Multi-cloud setup
- 5:03Our contributions
- 6:06Scheduler core
- 9:58Scheduler benefits
- 11:02Extending scheduler to edge
- 12:26Infrastructure challenges
- 15:36Proposed architecture
- 16:17Telemetry system
- 17:22Use cases
- 18:47Factory exoskeleton
- 19:29Drone application
- 20:15Q&A
related talks
39mIoT Application Running on KubeEdge + ARM Platform
40mDeep Dive: KubeEdge
41mKubeEdge
41mKubeEdge
01h 16mIntro + Deep Dive: Kubernetes IoT Edge WG
51mNo-compromise solutions for the integration between Edge and Cloud
33mIntro: KubeEdge
32mUsing k8s, Bare Metal, & 5G to achieve autonomous drone delivery!
