Lightning Talk: Scaling Time Series Databases on Kubernetes
Lightningat Data on Kubernetes Day North America 2023 · on Monday, 2023-11-06· at 16:55· W471ab, McCormick Place·
summary
Speedscale runs TimescaleDB to track billions of API interactions and claims single-digit-hundreds-millisecond queries at scale. Deploying the Helm chart works but persistent volumes cause multi-AZ downtime, so the cluster switches to Carpenter for zone-aware node provisioning and right-sizing. Moving the indexer from an external Lambda into the cluster yields 10x faster ingestion, and realistic traffic replay storms expose tuning needs. Result: high-ingest, queriable TSDB on Kubernetes that stays cheap and responsive (~150 ms typical).
AI-written summary · about
chapters (5)
- 0:00Introduction to time series
- 1:19Choosing TimescaleDB
- 2:13Solving storage challenges
- 4:20Networking and ingest
- 5:26Tuning and scaling
related talks
30mScaling Time-Series Data to Infinity: A Kubernetes-Powered Solution with Envoy
51mBuilding a scalable time-series database on PostgreSQL
07mLightning Talk: My Database Runs on Kubernetes. What's Next? Data Platforms!
10m⚡ Lightning Talk: Navigating the Complexities of Stateful Workloads in Kubernetes
06mLightning Talk
39mTimescaleDB: Re-engineering PostgreSQL as a time-series database
33mExponential Growth Without Sacrificing Platform Speed
13mTransform your business with time-series data (PRT061)