Why Old Data Becomes Expensive Over Time

Описание: How does old data affect your database over time?

In this video, we explore how aging data changes your database’s cost model — not just in terms of money, but in operational complexity, performance trade-offs, and maintenance overhead.

The real problem isn’t how much data you store.
It’s how that data behaves as it gets old.

Through two real-world examples — analytical workloads and logging systems — we examine:

Why old data is accessed differently than new data
How performance expectations change over time
Why maintenance burden increases as data grows
What happens when deletion patterns don’t match your database design

We’ll also discuss how different database systems encode different assumptions about data lifecycle — and why mismatches between those assumptions and your workload can lead to operational pain later.

Topics covered include:

Data lifecycle design
Operational cost of old data
Row-by-row deletion vs bulk deletion
Time-series workloads
Analytical queries and large scans
Database design trade-offs

If you’re working with relational databases, Elasticsearch, time-series systems, or designing backend architectures, this perspective will help you think more clearly about how your system ages over time.

Because database design isn’t just about features.
It’s about how your data evolves.

---

0:00 Introduction – Why Old Data Changes Database Cost
0:57 What Is a Database Cost Model? (Operational & Performance Impact)
1:48 The Real Problem: How Old Data Behaves Over Time
2:17 Access Patterns for Old vs New Data
3:06 Example - Analytical Workloads & Large Data Scans
3:59 Separating Fast Writes and Fast Reads (Ingestion vs Analytics)
7:19 Example - Systems Designed for Reliability
8:47 Elasticsearch Data Tiers Explained (Hot, Warm, Cold)
9:53 The Common Pattern Behind Both Examples
10:36 Why Old Data Becomes Operationally Expensive
12:14 Database Assumptions About Data Lifecycle
12:58 Real-World Logging System Example (Row Deletion vs Bulk Deletion)