Performance testing in the cloud
summary
Case study of an enterprise media platform over 3+ years shows cloud performance testing as practical engineering: tests run primarily on AWS, measure single‑instance breaking points (roughly 30 RPS in examples), tune autoscaling rules and monitoring, and prove cost tradeoffs between many small containers or fewer big ones. Disaster recovery is exercised across pilot light, warm standby and active-active patterns; autoscale reacts in minutes so downscale is deliberately slower than upscale; features like Spike arrest and CPU target ~50% become concrete knobs that require tuning to work.
AI-written summary · about
chapters (9)
- 0:00Cloud performance testing intro
- 2:20Why performance testing matters
- 4:11Cloud migration stages
- 8:56Resiliency and disaster recovery
- 16:25Performance engineering for cost
- 25:59Testing cloud scalability
- 41:27Optimizing cloud costs
- 50:05Performance testing as a service
- 1:03:09Q&A
related talks
57mLast- und Performancetests in der Cloud?!
48mTesting The Cloud
12mExponential Load Testing...
33mThe Real Cloud Testing Story
33mPerformance Testing in Agile and DevOps
43mBest practices for continuous testing and observability
46mChaos Experiments - Datacenter Stress Testing
31mPerformance Engineering and Testing of Apps and Services Deployed on Cloudfoundry