Scaling Your Cluster (Both Ways)
Talkat KubeCon + CloudNativeCon North America 2019 · on Wednesday, 2019-11-20· at 16:25· Room 6F - San Diego Convention Center, Upper Level·
summary
Shows how to scale pods and node capacity using virtual kubelet backed by Azure Container Instances (ACI) and how KEDA integrates with Kubernetes’ Horizontal Pod Autoscaler to scale workloads to and from zero. Demo runs on AKS, installs KEDA via Helm, pushes events into RabbitMQ and observes the HPA create up to 30 ACI-backed pods, then KEDA drives the deployment back to zero when the queue drains. Also contrasts using the cluster autoscaler (whole VMs) versus virtual kubelet (one pod per VM-like instance) and notes provider limits determine max scale and startup latency.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 2:00Virtual Machines
- 2:40Virtual Kubernetes
- 6:38Virtual Cubelet
- 8:05Scaling to Zero
- 8:52Kubernetes Centric
- 10:06Kara
- 10:52DeArchitecture
- 11:59Events Sources
- 12:32Demo Time
- 13:00Demo Setup
- 13:30RabbitMQ Deployment
- 13:58RabbitMQ Consumer Deployment
- 14:39RabbitMQ Publisher Deployment
- 16:03Cloud Native Application Bundle
- 17:09Demo
- 26:13Questions
related talks
41mEvent-driven autoscaling on Kubernetes with KEDA and Azure Functions
37mKubernetes Event-driven Autoscaling with KEDA
34mAutoscaling Made Easy with Kubernetes Event-Driven Autoscaling (KEDA)
07mAutoscaling with virtual node and Azure Kubernetes Service (AKS)
50mEvent-driven autoscaling on Kubernetes with KEDA and Azure Functions
21mMusic Matched to Different Ways to Scale in K8s, What Could Go Wrong?
35mKEDA - Real Time And Serverless Scaling In Kubernetes
38mHow Event Driven Autoscaling in Kubernetes Can Combat Climate Change