From Development to Production
Scale your TimescaleDB deployment
Module: Tiger Data 101 → An Introduction to TimescaleDB
Section: Putting TimescaleDB Into Action
Estimated time: 10–12 minutes total (this page: ~2 minutes)
Learning objectives
Section titled “Learning objectives”By the end of this module, you will be able to:
- Evaluate TimescaleDB capabilities against specific use cases
- Design solutions that scale from development to production
- Analyze real-world case studies and apply lessons learned
Development vs. Production
Section titled “Development vs. Production”Development
Section titled “Development”- Single machine TimescaleDB instance
- Standard PostgreSQL client libraries
- Manual data management (compression, tiering, retention)
Production (Tiger Cloud)
Section titled “Production (Tiger Cloud)”- Fully managed service (automated backups, scaling, monitoring)
- High availability with read replicas
- Built-in connection pooling
- Automated lifecycle policies
Connection Pooling
Section titled “Connection Pooling”Essential for applications with many concurrent connections:
Application (1000 connections) ↓PgBouncer (connection pool, 100 actual connections) ↓TimescaleDBBenefits: Fewer database resources, faster connection recycling
Schema Design Best Practices
Section titled “Schema Design Best Practices”-
Use appropriate time column types:
TIMESTAMPTZfor global applications (timezone-aware)TIMESTAMPfor local times
-
Denormalize selectively:
- Store device metadata alongside readings
- Reduces JOIN overhead
-
Index strategically:
- (time, tag) composite index for time-filtered, tagged queries
- (tag, time) for tag-filtered, time-ordered queries