Serverless on AKS (Azure Kubernetes Service)
summary
Shows how to run serverless on AKS using virtual nodes (ACI) so pods run on container instances, KEDA to convert queue depth into autoscaling and KEDA scales to zero, Dapr to provide platform-agnostic bindings, and OAM to separate developer app spec from infra. Demo puts 100 Service Bus messages on a queue, KEDA brings a deployment from 0 up to five pods, processes them, then scales back to 0 — it works. Notes: Azure Functions still outperforms if lock‑in is acceptable; AKS adds ops complexity.
AI-written summary · about
chapters (15)
- 0:00Introduction and background
- 1:06From monoliths to microservices
- 3:36Containers and Kubernetes basics
- 8:20Kubernetes scaling and autoscaling
- 11:26Synchronous vs asynchronous services
- 15:57Serverless functions and FaaS
- 19:39Kubernetes vs serverless trade-offs
- 23:42Running serverless on Kubernetes
- 27:11Virtual nodes and KEDA scaling
- 30:03Demo: KEDA with service bus queue
- 37:11Dapr for event-driven bindings
- 41:10Open Application Model overview
- 45:27Demo: Dapr and OAM deployment
- 49:46When to use serverless vs Kubernetes
- 52:15Q&A and common Kubernetes challenges
related talks
46mModernizing your applications with containers and serverless
01h 01mDevOps with Azure Kubernetes Service
41mEvent-driven autoscaling on Kubernetes with KEDA and Azure Functions
29mServerless on Kubernetes: What We Learned From Customers
57mAzure Kubernetes Service (AKS) for Beginners
50mEvent-driven autoscaling on Kubernetes with KEDA and Azure Functions
31mScaling Your Cluster (Both Ways)
47mIntegrating Azure AI and Azure Kubernetes Service to build intelligent apps