Health checks & dependency management
summary
A web recommendation app runs on EC2 in an Auto Scaling group across AZs, uses Amazon DynamoDB for data, and implements a Python /healthcheck path that returns 200 or 503. Health checks call the external recommendation API (hard vs soft dependency) and can fail an instance closed so the load balancer removes it; if every target fails the LB can fail-open and route anyway. Soft fallback returns a static suggestion (I Love Lucy) so the site stays usable.
AI-written summary · about
chapters (10)
- 0:00Health checks & dependency management
- 3:50Infrastructure reliability
- 6:16Basic health checks
- 7:39Deep health checks
- 9:43Fail closed vs fail open
- 11:49Hard to soft dependencies
- 13:36Resilient components
- 15:21Automated mitigation
- 16:02Unexpected failures
- 17:25Resources & call to action
related talks
31mDevOps in the Cloud: Case Studies of Amazon.com teams and their resilient architectures
21mAWS On Air ft. Amazon SageMaker Inference Recommender
26mA proactive approach to resilience in modern apps
46mWhat You'll Miss on AWS & How To Find It
41mAutoScaling Web Application Security in the Cloud
42mPlanning To Fail
42mChaos Engineering: Why Breaking Things Should be Practiced
28mApplication Health Checks Done the Right Way