
- Skim it in one sitting, or
- Bookmark it and come back when a term shows up in a meeting or an interview.
How to use this guide
- Skim (10 minutes): read the headings + first sentence.
- Study (30–45 minutes): pause on anything unfamiliar and search one example.
- Interview drill (5 minutes): explain 5 concepts out loud (definition → trade‑off → example).
A) Core trade-offs (10)
1) Vertical vs horizontal scaling
- Vertical: bigger machine.
- Horizontal: more machines.
- Horizontal wins long-term, but forces you to handle coordination, statelessness, and failure.
2) Latency vs throughput
- Latency: how fast one request finishes.
- Throughput: how many requests per second.
- You can increase throughput and still make users miserable if tail latency is bad.
3) Tail latency (p95/p99)
4) CAP theorem
5) PACELC
6) ACID vs BASE
- ACID: strict transactions, correctness first.
- BASE: availability + eventual consistency, correctness converges over time.
- Many real systems use both: money = ACID, feeds/analytics = BASE.
7) Stateful vs stateless services
8) Coupling and boundaries
9) Idempotency
10) Backpressure
B) Networking and communication (10)
11) Load balancing
12) Load balancing strategies
13) Reverse proxy vs forward proxy
- Reverse proxy: represents servers.
- Forward proxy: represents clients.
14) API gateway
15) Service discovery
16) Timeouts and retries
17) Circuit breaker
18) HTTP/2 and HTTP/3
19) gRPC vs REST
- REST: simple and universal.
- gRPC: typed contracts, streaming, efficient service-to-service.
20) WebSockets vs SSE
- WebSocket: two-way realtime.
- SSE: server-to-client stream (simpler).
C) Data storage and modeling (10)
21) Choosing the right primary database
22) Indexes (B-Tree vs LSM)
- B-Tree: great for reads and ranges.
- LSM: great write throughput; reads are more work.
23) Write-ahead logging (WAL)
24) Normalization vs denormalization
25) Sharding (partitioning)
26) Re-sharding
27) Replication
28) Strong vs eventual consistency
29) Consistent hashing
30) Polyglot persistence
D) Caching and performance (10)
31) Cache-aside
32) Read-through / write-through
33) TTL and invalidation
34) Cache stampede
35) Hot keys
36) CDN caching
37) Pagination vs infinite scroll
- Pagination is simpler for stability and caching.
- Infinite scroll is nicer UX but can be expensive and tricky.
38) Data compression
39) Batching
40) Async processing
E) Reliability, operations, and security (10)
41) Message queues
42) At-least-once vs at-most-once vs exactly-once
43) Dead-letter queues (DLQ)
44) Distributed locks
45) Leader election
46) Observability
47) SLOs and error budgets
48) Rate limiting
49) Authentication vs authorization
- AuthN: who are you?
- AuthZ: what are you allowed to do?
50) Threat modeling
A quick interview drill (5 minutes)
- Definition (one sentence)
- Why it exists (the pain it solves)
- Trade-offs (what you pay)
- Example (one real system)