MySQL Cluster as Transactional NoSQL
summary
MySQL Cluster claims to combine full ACID SQL with native key‑value NoSQL access, offering ACID transactions alongside Memcached API and Java/Node/C++ clients; it runs in‑memory with online schema/config changes and transparent auto‑sharding for linear horizontal scale. The demo numbers show up to 20 million updates/sec on 30 data nodes and comfortably over a million on smaller clusters. It supports active‑active clusters with asynchronous replication to other clusters or MySQL servers and advertises 99.999% availability.
AI-written summary · about
chapters (8)
- 0:00MySQL performance tools
- 3:41MySQL Cluster as NoSQL
- 10:55Why NoSQL in MySQL Cluster
- 15:40Technical implementation
- 20:47NoSQL APIs
- 24:32Replication and storage
- 28:47Cluster and server NoSQL
- 32:04Future NoSQL developments
related talks
59mMySQL (NDB) Cluster Best Practices
29mModern Database Architecture: From SQL to NoSQL & Distributed SQL
55mThe Future of MySQL
58mDon’t Drop ACID - Transactions in Distributed NoSQL
25mThe future of databases: Combining operational and analytical data at massive scale
27mDistributed MySQL Architectures: Past, Present and Future
01h 32mEffective Design Patterns in NewSQL
58mDecentralized Cluster Nodes