Improving the Reliability of Cloud Foundry Applications Through Advanced Health Checks
Talkat Cloud Foundry Day North America 2024 · on Wednesday, 2024-05-15· at 14:45· Education Center, 2nd Floor·
summary
Cloud Foundry adds configurable startup, liveness, readiness checks plus health-check types HTTP, Port, Process (Process default). Startup failures leave the instance for developer action, liveness failures trigger restarts with a three‑attempt then exponential backoff up to ~200 attempts, and readiness failures remove instances from routing without restarting them. Manifest keys (liveness_/readiness_) let apps use HTTP endpoints or TCP checks and tune timeouts and intervals. Team also adds AZ-aware deployment and cross-component distributed tracing; a recent AWS AZ outage is recovered by redeploying to other AZs in about two hours.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:51About the speaker
- 2:33Cloud Foundry objectives
- 7:45Handling infrastructure failures
- 9:38Why improve health checks
- 11:03Community-driven improvements
- 11:46Types of health checks
- 16:06Crash event handling
- 17:12Health check configurations
- 18:36Health check use cases
- 19:40Additional reliability tools
related talks
13mLightning Talk: Cloud Foundry Monitoring: How to Never Leave Your Deployment Unattended
40mCloud Foundry Availability
36mCloudy With a Chance Of Chaos: Verifying the Resiliency Of Cloud-Native Applications
20mAdvancing the Future of Cloud Foundry
01mCloud Foundry: The Right Tool for the Right Job
32mBuilding a Robust Cloud Foundry (HA, Security and DR)
33mCloud Foundry Weekly: Live at CF Day Germany
55mChaos Engineering for Cloud Foundry Platform