The System Design
Roadmap.

Round Robin, Weighted RR, Least Connections, IP Hash — how traffic gets spread across servers.

Cache capacity, eviction policy (LRU/LFU/FIFO), and how hit rate changes latency.

1NF → 2NF → 3NF with real tables, and the anomalies normalization actually fixes.

Full table scan vs B-Tree traversal — the O(n) vs O(log n) difference, visually.

Sync/Async/Semi-Sync replicas, replication lag, and failover.

Hash, Range, or Directory sharding across shards — and what makes a shard "hot".

A virtual ring of servers where adding/removing one only remaps a minimal set of keys.

Inject a network partition — watch CP systems reject reads while AP systems serve stale data.

Kafka vs RabbitMQ — producers, partitions, consumer groups, and lag.

Publish to a bus, watch subscribers react in parallel — vs request-response, side by side.

Services competing for a Redis lock with TTL — crash the holder, watch it transfer.

A saga across Order → Payment → Inventory → Shipping, with automatic compensation on failure.

Token Bucket, Leaky Bucket, Fixed/Sliding Window — with a live 429 rejection counter.

No Retry vs Fixed vs Exponential vs Jitter — watch a thundering herd collapse into a smooth curve.

Closed → Open → Half-Open, live, as failure rate crosses your threshold.

Duplicate payment requests, with and without an idempotency key, side by side.

Auth, rate limiting, caching, transform — one pipeline, inject failures to see 401/429/503.

Scale instances, kill one mid-flight, watch the registry route around the failure.

JWT, OAuth 2.0 Authorization Code, and Client Credentials — decode a token live.

Inject error/latency spikes and memory leaks — watch alerts, logs, and traces respond.