Topic S.5
The Vocabulary: Reading Any Design Diagram
In one line
Fifteen words cover most of what you'll hear in design discussions: latency, throughput, availability, consistency, scalability, SPOF, replication, sharding, cache, queue, idempotency, and more. Learn them once with plain definitions.
Think of it like this
Learning a sport's rules before watching a match. Once you know 'offside' and 'penalty', commentary makes sense. These words are the rules of system design commentary.
Key ideas
- 01
LATENCY: how long one request takes (ms). THROUGHPUT: how many requests per second the system handles. You can have high throughput and bad latency (a busy but slow checkout line), and improving one can hurt the other.
- 02
AVAILABILITY: the share of time the system works, written as 'nines': 99.9% ≈ 8.8 hours down per year, 99.99% ≈ 53 minutes. RELIABILITY / DURABILITY: the system does the right thing and doesn't lose data. SPOF (single point of failure): one part whose failure takes everything down.
- 03
SCALABILITY: handling more load by adding resources. VERTICAL scaling = a bigger machine; HORIZONTAL scaling = more machines. STATELESS: a server keeps no per-user data between requests, so any server can handle any request.
- 04
CONSISTENCY: whether everyone sees the same, latest data. STRONG consistency: reads always see the newest write. EVENTUAL consistency: copies catch up after a short delay. REPLICATION: keeping copies of data on several machines. SHARDING (partitioning): splitting data across machines by a key.
- 05
CACHE: fast copy of data to avoid slow trips. QUEUE: a buffer where producers leave messages for consumers to process later. IDEMPOTENT: doing it twice has the same effect as doing it once (safe to retry). BACKPRESSURE: slowing producers when consumers can't keep up. SLA/SLO: the promised/targeted level of service.
Code & diagrams
Availability Downtime per year Downtime per month
99% 3.65 days 7.3 hours
99.9% 8.8 hours 43.8 minutes
99.95% 4.4 hours 21.9 minutes
99.99% 52.6 minutes 4.4 minutes
99.999% 5.3 minutes 26 secondsExplain without notes
Explain latency vs throughput with a real-world example.
What's the difference between replication and sharding?
Practice
Label each part of the Topic S.4 diagram with the vocabulary words that apply (e.g. which parts are replicated, which are stateless, where the SPOFs are).
Run it in production
You've designed it. Now build, operate, and break the same idea hands-on in the DevOps courses:
Completion checklist
I can define every word in this topic without notes
I can annotate a design diagram with these terms