User Space in Cloud Platforms
summary
Heroku explains how apps run down the stack from MRI Ruby through LXC containers on Ubuntu into cgroups and namespaces and then into Amazon VMs using the paravirtual API, showing that the platform builds an API to start and stop containers and enforces resource limits. The concrete tradeoffs are visibility loss into the kernel, measurable operational efficiency gains from multiplexing CPUs, and the pragmatic choice to self-host services when kernel boundaries prove too limiting.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:58Kernels User Spaces
- 2:38Ruby Virtual Machine
- 3:18Heroku
- 4:10LXE
- 5:58Zen
- 6:53Amazon
- 8:53Physics
- 10:55Visibility
- 12:02Operational Efficiency
- 13:51expressiveness
- 14:50reduce complexity
related talks
23mCgroups. Using container technologies as a PaaS provider
47mContainers - the little Linux feature revolutionizing apps
34mLook ma, no OS! Unikernels and their applications
27mServerless MicroVMs for Containerized Applications
41mContainers or Serverless Functions - When to choose what?
31mServerless Rails: Understanding the pros and cons
25mPreparing Apps for Dynamic Scaling
01h 01mYou have a container image: Now what?