Concept: Reliability

1. The Snapshot

A system's ability to continue working correctly (performing the correct function at the desired performance level) even when things go wrong.

2. The Description

Reliability is the bedrock of data systems. It is not about preventing all errors (which is impossible) but about designing systems that tolerate "faults" so they don't escalate into total system "failures." It combines correctness, resilience, and security.

3. Author Quotes

"If all those things together mean 'working correctly,' then we can understand reliability as meaning, roughly, 'continuing to work correctly, even when things go wrong.'" (p. 6)

4. Defining Features

• Correctness: The application performs the expected function.
• Resilience: The system tolerates user mistakes or unexpected usage.
• Performance: Remains "good enough" under expected load/volume.
• Security: Prevents unauthorized access or abuse.

5. The Boundary

• It is NOT Fault-Proof: Reliability doesn't mean faults never happen; it means the system is designed to handle them so they don't lead to a total outage.
• It is NOT just Availability: A system can be available (accepting requests) but unreliable (returning wrong data).

6. The Prototype

An online banking system. If a single database node crashes (fault), the system automatically fails over to a replica so the user transaction still completes successfully (reliable).

7. Helpful Info

Reliability is relative to expectations. A prototype might be "reliable" if it crashes once a week, but a pacemaker must be reliable for years. The goal is to define the boundary of what "working correctly" means for your specific domain.

8. The Swap Test

"High Reliability requires us to anticipate potential Faults and design the system to be Fault-Tolerant."

9. Source Reference

ddia/pages/page_028.txt

---

🧠 Pedagogical Tracking

Milestone	Status	Date	Lesson Ref	Notes
Introduced in Lesson	✅	2026-01-28	Lesson 01	Initial Study
Active Recall #1	⚪
1-Day Review	⚪
1-Week Review	⚪
1-Month Review	⚪