Bret Piatt interviewed at Velocity 2010
summary
Cassandra forgoes a fixed relational schema in favor of columns and column families, returning slices of rows that applications must stitch together rather than joins; it fits high‑volume use cases at Twitter, Digg and Facebook and pays off when data grows into terabytes. Recommended minimum deployment is roughly an eight node cluster with about 25 GB per node (a few hundred GB total) before switching from relational systems. Open standards reduce lock‑in, and cloud economics push teams to put your application next to the data.
AI-written summary · about
chapters (5)
- 0:00How does Cassandra differ from relational database
- 0:45What types of sites and services use Cassandra
- 1:37Tipping point for moving to Cassandra
- 2:31Why are open standards important
- 3:56Cloud computing
related talks
29mCassandra - an introduction
35mThe What and Why of Cassandra Datamodelling at Scale
34mBig and Small Data at @Twitter
25mData in the Cloud
01h 05mRoundtable: Cloud Databases as a Service
06mProject Lightning Talk: Vitess in Production: Powering Applications at Massive Sca...
27mCloning Twitter: Rails + Cassandra = Scalable Sharing
34mData Storage: NoSQL Toasters and a Cloud of Kitchen Sinks