Edge-native applications - what happened to cloud-native?
summary
Claims that server-side WebAssembly lets one build once and run everywhere with concrete tooling: Spin CLI, server-side WebAssembly, and selective deployments. Demo compiles multi-language app to a ~40 KB .wasm, idles at <1m CPU and 12–20 MB RAM, and handles millions of short-lived executions; Kubernetes runs modules via Spin Cube using a containerd shim and scales to 20 replicas in seconds. Edge example: Vision Pro needs 14 ms end-to-end on-device but 1 ms from a nearby edge POP. OCI artifact push (ttl.sh) and Firmian Wasm Functions distribution complete the pipeline.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:52Cloud-native definition
- 5:26Cloud-native vs serverless
- 8:18Edge and compute continuum
- 12:12WebAssembly for portable compute
- 14:00Selective deployments concept
- 15:48Spin framework
- 22:18Spin apps locally
- 23:00Spin apps on Kubernetes
- 27:18Multi-component app demo
- 30:09Local dev workflow demo
- 39:14Component deployment control demo
- 41:47Edge deployment with Firmian
- 48:34Key takeaways
- 55:00Q&A
related talks
24mA Greener, Cost Effective Cloud with Serverless WebAssembly
44mA Greener, Cost-Effective Cloud with Serverless WebAssembly
41mRevolutionizing Cloud Native: WebAssembly's Role in Serverless...
30mThe Future of the Cloud is WebAssembly
22mWebAssembly beyond the browser
36mA Greener, Cost-Effective Cloud with Serverless WebAssembly
59mNext generation microservices and serverless apps with WebAssembly & Spin
30mA Greener, Cost-Effective Cloud with Serverless WebAssembly