DSS Keynote — Back to the Future: Prepare Your Data Infrastructure for Anything
summary
Claims that cloud-native deployments must handle real failures and global latency with concrete numbers: read replicas cut reads to ~4–5 ms, geo-partitioned clusters cut writes to ~8 ms and an RF3 cluster survives a full zone outage with no app impact. Cited industry data shows seven major cloud-region failures last year (about one every 51 days). Temenos benchmark targets ~500k DB transactions/sec for a 100M-user bank. CDC and managed tools let change streams and debugging go from ~10 weeks to a single day, and a slow‑query fix yields a 4× speedup.
AI-written summary · about
chapters (8)
- 0:00Welcome and keynote theme
- 2:50Embracing cloud-native
- 11:47Protecting against failures
- 18:37Data security and compliance
- 21:05Handling workload uncertainty
- 30:32Multi-tenant database scale
- 41:14Developer productivity
- 50:32Ongoing performance tuning
related talks
40mLatency, Throughput & Fault Tolerance: Building a Modern Database
57mNo Downtime, No Limits
48mResiliency and Availability Design Patterns for the Cloud
31mReimagining the Future: Where the Market is Evolving
31mFuture-proofing operational data at scale
01h 05mRoundtable: Cloud Databases as a Service
40mDistributed Databases: What Could Possibly Go Wrong?
11mChallenges Deploying Modern Real-Time Payment Systems on Traditional RDBMS