Glossary

Capacity

Answers question about how much do I have to host in a system in a given time? This question can be answered for persistent storage as well as in-memory storage.

Data Integrity

Answers questions like:

• is all the data being written to storage?
• is all the data being read from storage?
• is all the data safe in the storage?

Fault Tolerance

Answers question - can I handle something when failures take place

Observability

A term used to monitor health and performance of a system. It tells you when the metrics represents a problem or is about to be one. Observability alerts you on such events, makes you aware of the business impacts, the indicator(s) of impact, and also guides you towards how to fix the problem.

Throughput and Latency

Throughput - explains how many requests can be processed and completed, by a system in any given amount of time (usually measured per second). Requests can be in terms of API request, Database queries, or even messages in a queue.

Latency - explains how fast those requests can be processed in a given amount of time (usually measured per second).

References

• Latency vs Throughput
  • YT - Latency vs Throughput | System Design Essentials
  • YT - Latency vs Throughput
  • What is the Difference Between Bandwidth and Latency?
  • Latency vs Throughput – Understanding the Difference
  • Network Latency vs. Throughput vs. Bandwidth
  • Bandwidth vs. Latency: What is the Difference?
  • Throughput Vs. Latency – What’s the Difference?
  • What Is Network Latency?
  • Understanding Network Bandwidth vs Latency