Microservices and Containers. How much faster than a VM?!
summary
Claim: containers instantiate in seconds versus minutes so infrastructure can reuse capacity in real time instead of autoscaling. Experiment: a simple scheduler experiment driving ECS Marathon Swarm Kubernetes and the Docker API on bare metal EC2 and Azure. Result: standard orchestrators and hardware keep up, making Google/Netflix-style reuse plausible. Surprise: measured potential utilization jumps from typical 10–15% to about 50–70%, cutting hosting costs and changing operational tradeoffs.
AI-written summary · about
chapters (9)
- 0:00Intro
- 3:06VM vs MS
- 6:00Containers
- 9:36Containers vs VMs
- 15:00Is it true
- 16:30The scheduler
- 17:30Is this possible
- 19:20What are they doing
- 21:20The Throne
related talks
23mMicroservices and containers
10mThe differences between microservices and containers
36mFrom VMs to Containers: Decompose and Migrate Old Applications
01h 04mDeploying and Scaling Microservices
41mOptimizing Microservices with WebAssembly and Linux Containers: Exploring Hybrid Contain...
39mMicroservices & Kubernetes in a High-demand Environment: The Experience
51mDeploying And Scaling Microservices
08mMicroservices at Scale