Apache Cassandra: What’s New and What’s Next
summary
Apache Cassandra now adds concrete features: Accord leaderless Paxos implements cross-partition transactions with strict serializability and usually a single-roundtrip latency, validated by a formal proof and millions-of-transaction simulations; transactional cluster metadata replaces fragile gossip to make schema and membership changes atomic and enable efficient range splitting; storage-attached indexes (SEI) in Cassandra 5 provide partition-local, zero-copy streaming indexes; witness replicas cut durable copies and promise roughly 30% storage savings while preserving quorum consistency for reads and writes.
AI-written summary · about
chapters (11)
- 0:00Cassandra's evolution
- 3:31Dynamo's design origins
- 7:22Scalability and availability
- 10:10Database-like features journey
- 14:24Quality challenges
- 16:53Transactional metadata SEI
- 19:17Witness replicas savings
- 20:20Distributed transactions Accord
- 23:52Future with distributed transactions
- 25:58Summary achievements
- 26:39Q&A identity
related talks
54mEvolving Transactions in Apache Cassandra
10mKeynote: Introducing Apache Cassandra® 5!
53mApache Cassandra. Inception - all you need to know
46mData Replication with Apache Accord
29mCassandra - an introduction
46mApache Cassandra: 8 years in production
45mCassandra Architecture: Understanding, Scaling and Optimizing
34mThe Accord Consensus Protocol: ACID transactions globally at any scale