Dystopia as a Service
summary
Claims a cloud-native stack built on AWS, Cassandra and S3 handles real traffic: 2 PB in S3, 100+ billion events per day, ~1/3 of North American fixed-peak video traffic via Open Connect CDN. Architecture uses ~400 services and 50+ Cassandra clusters (largest 144 nodes) with 3× replication on ephemeral disks; an 18 TB restore and cross-country copy runs at ~9 Gbps with ~0.5 s eventual consistency. Chaos engineering runs like Chaos Gorilla (kill one-third of infra) surface faults but quorum keeps service running. Tools: Astyanax, Priam, Genie, Zuul, Denominator.
AI-written summary · about
chapters (24)
- 0:00Introduction
- 1:14Cloud Native
- 5:00Iterating
- 8:50Inspiration
- 9:47Thinking in Systems
- 10:50Antifragile
- 12:43Drift into Failure
- 15:00Obvious vs Right
- 15:42Service Types
- 17:24Cloudnomics
- 18:51Transitions
- 22:36Management
- 28:14Continuous Delivery
- 30:50Netflix
- 35:04Storage Tier
- 39:07tomcat
- 42:44Big Data
- 48:54Global Architecture
- 51:09Cross Region
- 51:24Benchmark
- 53:39Traffic Routing
- 55:52Incidents
- 58:07Cloud Prize
- 58:43Cloudnomics Prize
related talks
41mDystopia as a Service
54mMigrating to Cloud Native with Microservices
27mScalable by Design: How to Build an Elastic Cloud Infrastructure...
31mDevOps in the Cloud: Case Studies of Amazon.com teams and their resilient architectures
28mCloud The Final Destination
16mBuild for real time on AWS: Thrive in an unpredictable market
48mResiliency and Availability Design Patterns for the Cloud
38mCloudy with a Chance of Chaos