Enabling Hypercore & Best Practices
Apply compression and manage data lifecycle
Tiger Data 101 → TimescaleDB | Section: Hypercore | ⏱ Time: ~3 min
Hot vs. cold data: mixed chunk access
Section titled “Hot vs. cold data: mixed chunk access”TimescaleDB allows you to keep recent chunks uncompressed (hot) while compressing older chunks (cold), all in the same hypertable:
sensor_readings├── Jan 2025 (compressed, 50MB) ← cold, compressed, queryable├── Feb 2025 (compressed, 48MB)├── Mar 2025 (compressed, 52MB)└── Apr 2025 (uncompressed, 500MB) ← hot, fast writes, recent dataBenefits:
- Recent data stays fast for writes and quick queries
- Old data compressed for storage savings
- No schema changes needed—hypertable hides the complexity
When to use hypercore
Section titled “When to use hypercore”Use compression when:
- Storing historical data you query infrequently
- Data has patterns (sensors, metrics, events)
- You need to reduce storage costs significantly
- Analytical queries outweigh random access
Avoid compression when:
- Data is frequently updated or deleted
- You need ultra-low latency random access
- Compression would add overhead that doesn't save space
Knowledge check
Section titled “Knowledge check”Knowledge Check
Question 1 of 3