Understand the core challenges of distributed computing — fallacies, time/order, failure models, and the fundamental differences from single-machine systems.

Master the CAP theorem's consistency-availability-partition tolerance trade-off, plus strong vs eventual consistency models used in real systems.

Learn how distributed systems agree on a single value despite failures — Raft's understandable approach and Paxos's elegant foundation.

The structural primitives that hold distributed systems together — consistent hashing with virtual nodes, Merkle trees, Gossip/SWIM, and the data structures behind ring-shaped coordination.

Explore asynchronous communication patterns — publish/subscribe, message brokers, stream processing, and exactly-once semantics.

Design and reason about microservice systems — service decomposition, communication patterns, service discovery, and observability at scale.

The distributed-systems view of caching — multi-level hierarchies, stampede prevention, dogpile/bounding, Bloom filters as cache guards, and CDN invalidation as a hard problem.

Coordinate state across services without a single ACID database — two-phase commit, the saga pattern, idempotency, and outbox tables in the real world.

Traffic shaping under overload — token buckets, adaptive concurrency, timeout budgets, hedged requests, and the laws that govern graceful degradation.

The engineering craft of designing systems at scale — back-of-envelope math, capacity planning, trade-off analysis, and worked designs for canonical problems like rate limiters, URL shorteners, and chat services.