Cloud Foundry AutoScaler Service: Features, Design and Roadmap
summary
Cloud Foundry AutoScaler is a service that implements dynamic scaling and scheduled scaling with concrete policy files (min/max instance limits, scaling rules, stat window and cooldown). It collects memory utilization, throughput and response-time metrics via a metric collector, aggregates them, and hands events to a scaling engine; components run as microservices backed by Postgres and can be deployed with BOSH or as CF apps. Beta runs on SAP, AWS and OpenStack, tests aim for ~2k apps.
AI-written summary · about
chapters (12)
- 0:00Introduction and overview
- 2:14Auto scaling basics and types
- 5:49Dynamic scaling policies explained
- 9:02Policy definition and scheduling
- 12:45End-user implementation steps
- 13:48Service provider deployment demo
- 16:18Architecture and components
- 18:29Production readiness and testing
- 20:43Future features and roadmap
- 22:32Microservices and moving parts
- 23:58Open connections issue
- 25:43Q&A and closing remarks
related talks
16mAutoscaling Distributed System with BOSH
33mCloud Foundry Weekly: Live at CF Day Germany
48mCloud Foundry - your PaaS on OpenStack
29mCustom App Autoscaler Using Cloud Controller API - Christopher Decelles
41mSpeed Up Your Scaling with Cloud Foundations
31mPut Your Money Where Your CF Is
15mThe Cloud Foundry Renaissance
21mCloud Foundry on Google Cloud Platform