Exponential Load Testing...
summary
Argues that moving from single-cloud to a multi‑cloud load testing platform produces clearer, more reliable results by generating traffic from AWS, Rackspace, Verizon, and Google Compute and correlating outcomes across providers. Finds that errors appearing in only one cloud are likely external, while errors across clouds indicate internal faults; measures show 5–45 minute spin‑up variability and frequent connection resets and socket timeouts. Implements orchestration with OpenStack or RightScale, benchmarks each provider, and warns about cross‑cloud transfer costs despite offering near-infinite capacity (practical limits).
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:00Information Technology vs Business Technology
- 1:43Traditional Load Testing
- 2:35Cloudbased Load Testing
- 5:23MultiCloudbased Load Testing
- 5:49MultiCloud Load Testing
- 7:24Relationship with the Service Provider
- 8:45Optimize for the Cloud
- 10:58Summary
related talks
57mLast- und Performancetests in der Cloud?!
01h 08mPerformance testing in the cloud
16m"Performance Testing: ..."
48mMulti-Cloud Magic: Leveraging Multi-Cloud Clusters in Real-World Scenarios
42mStorage Performance Testing in the Cloud
45mMultiCloud
46mDown-to-Earth Cloud Scaling
34mLessons Learned from Five Years of Multi-Cloud