Choosing the Right Storage Solution

Tiger Data 101 → TimescaleDB | Section: Working with Time-Series Data | ⏱ Time: ~2 min

When You Need a Time-Series Database

You should consider a specialized time-series database when:

✅ Consider Time-Series Storage If...

• High write volume: Thousands to millions of writes per second
• Append-only pattern: Data rarely or never updates once written
• Large dataset: Terabytes to petabytes of data
• Time-range queries: Most queries filter by time ranges, not entity IDs
• Retention policies: You need automatic data lifecycle management (compression, downsampling, deletion)
• Real-time analytics: You need fast aggregations over recent data

❌ Stick with Relational Storage If...

• Small dataset: Gigabytes to low terabytes
• Frequent updates: Data changes often after initial write
• Entity lookups: Most queries search by ID or key, not time ranges
• Complex transactions: You need ACID guarantees across multiple tables
• Relational structure: Your data is highly normalized with many foreign keys

Note

Remember: TimescaleDB is a PostgreSQL extension, so you can use both storage patterns in the same database. Relational tables for your user and order data, hypertables for your metrics and events.

---

The TimescaleDB Advantage

TimescaleDB is purpose-built for time-series data:

Aspect	Standard PostgreSQL	TimescaleDB
Write throughput	Good for moderate load	Optimized for millions/sec
Storage	Row-based	Row and columnar (compression)
Time-range queries	Slower (full-table scans)	Fast (time-partitioned chunks)
Lifecycle management	Manual	Automatic policies
Compression	Not built-in	Native columnstore compression
Aggregations	Resource-intensive	Continuous aggregates

---

Decision Tree

Here's a simple decision tree to help you choose:

Do you have time-ordered data with high write volume?
├─ NO  → Use standard PostgreSQL/MySQL
└─ YES
   └─ Is it mostly append-only?
      ├─ NO  → Use standard PostgreSQL/MySQL
      └─ YES
         └─ Do you need automatic lifecycle management?
            ├─ NO  → Time-series storage can help, but not critical
            └─ YES → Use TimescaleDB or a specialized TSDB

---

Next: Build on This Foundation

You've now mastered the fundamentals of time-series data! Next, you'll learn how TimescaleDB is built and the three core features that make it powerful:

• Hypertables: Automatic time-based partitioning
• Hypercore: Columnar storage and compression
• Continuous Aggregates: Pre-computed rollups for fast queries

---

Knowledge check

Test your understanding of time-series data concepts with this quiz. You'll get immediate feedback and resource recommendations for any questions you miss.

Knowledge Check

Question 1 of 4

A. Which of the following is NOT a characteristic of time-series data?

A)A) Data points are ordered by timestampB)B) Data is rarely updated once writtenC)C) Queries most often filter by primary key lookupD)D) Write volume is typically high and continuous

Check Answer

---

NEXT CHAPTER→

Hypertables

Learn how TimescaleDB automatically partitions time-series data into chunks and enables efficient querying and management at scale.