Running Netflix on Cassandra in the Cloud

Explore Netflix's transition from a traditional datacenter-based architecture to a large-scale cloud-based system using Cassandra in this conference talk. Dive into the automation and open-sourcing of Cassandra management for multiple clusters and hundreds of nodes using Priam and Astyanax. Learn about backups, archiving, and performance and scalability benchmarks. Discover the Cassandra service pattern, production deployment strategies, high availability techniques, and triple replicated persistence. Examine Cassandra instance architecture, query examples, and data flow optimizations for multi-region writes. Gain insights into Cassandra backup processes, the Cassandra Explorer tool, and ETL for Cassandra. Review open-source projects, scalability tests on AWS, and performance comparisons between disk and SSD. Witness a live demo workload and Jmeter setup, and explore future steps, including the "Hacktion" concept and implementation plans.

Introduction
Cassandra Service Pattern
Production Deployment
High Availability
Triple Replicated Persistence
Cassandra Instance Architecture
Priam - Cassandra Automation
Astyanax Query Example
"Traditional" Cassandra Write Data Flows
Astyanax - Cassandra Write Data Flows
Data Flows for Multi-Region Writes Token Aware, Consistency Level - Local Quorum
Cassandra Backup
Cassandra Explorer
ETL for Cassandra
Open Source Projects
Scalability from 48 to 288 nodes on AWS
Cassandra on AWS
Cassandra Disk vs. SSD Benchmark
Live Demo Workload
Jmeter Setup
Next Steps
The Hacktion
Implementation plan