Part 1 of 9 · 1 min read
Why this page comes first
In this module you will
- Define reliability, scalability and maintainability, and give a plain example of each.
- Tell a fault from a failure, and name the three common sources of faults: hardware, software and people.
- Describe load with a few numbers (load parameters) and explain why the right numbers depend on the system.
- Read response-time percentiles (p50, p95, p99) and say why the average hides the slow requests.
- Explain tail latency amplification with a worked calculation.
- Name the three parts of maintainability: operability, simplicity and evolvability.
Every design in this path is judged against three goals. Many later words, such as replication (keeping copies of data), caching (keeping fast copies of answers), queues (waiting lines for jobs) and monitoring (watching the system), exist to serve one of them. This page names the three goals and the vocabulary used to measure them.
The ideas here come from the first chapter of the book "Designing Data-Intensive Applications". Later pages return to them many times.