Essential Architecture Exam Topics: A Must‑Read Review Guide

1. Must‑know topics : Master the five classic architecture styles—pipeline‑filter, broker, event‑driven, layered, and C/S vs. B/S—knowing their definitions, pros/cons, applicable scenarios, and impact on quality attributes. For quality‑attribute questions, memorize the six elements (stimulus source, stimulus, environment, artifact, response, response measure) and the six attribute types (availability, modifiability, performance, security, testability, usability), with concrete targets such as 99.99% availability, recovery time under 30 seconds, and response time below 200 ms for 10 k concurrent users.

2. ATAM evaluation : Remember the nine ATAM steps and the four analysis outcomes (sensitivity, trade‑off, risk, non‑risk). Build utility trees as required, and differentiate ATAM from CBAM, which adds cost analysis.

3. Microservice architecture : Compare microservices with SOA and monolithic designs across coupling, deployment, data management, and team organization. Key components include service registry/discovery, API gateway, configuration center, circuit breaker, tracing, and load balancing. In papers, justify service splitting using DDD bounded contexts and illustrate benefits such as a 90% reduction in system‑wide failure rate after introducing circuit breaking.

4. Big‑data architectures : Contrast Lambda (batch + speed + serving layers) with Kappa (stream‑only), noting Lambda’s strong fault tolerance versus Kappa’s simplicity and higher stream‑processing demands. Recognize Flink as the core stream‑batch integration engine. Quantify gains when migrating from Lambda to Kappa, e.g., data‑processing latency dropping from hours to seconds and operational cost decreasing by 40%.

5. Cloud‑native fundamentals : Understand the five characteristics—containerization, microservices, DevOps, continuous delivery, declarative APIs. Master Kubernetes orchestration, service mesh, and serverless scenarios. Example: scaling a Kubernetes deployment raises QPS from 500 to 5 000 and improves resource utilization by 35% during peak traffic.

6. High‑concurrency web architecture : Design layered solutions using CDN, load balancer, application tier, cache layer, and database tier. Quantify impact, such as Redis caching reducing DB QPS from 8 000 to 800 and latency from 800 ms to 120 ms.

7. Distributed systems : Apply CAP and BASE theories, and choose appropriate transaction models (2PC, 3PC, TCC, Saga) based on scenario—e.g., financial transfers require 2PC, while e‑commerce flash sales favor TCC or Saga. Include Redis and ZooKeeper implementations for distributed locks.

8. Domain‑Driven Design (DDD) : Use bounded contexts to split services into domains (order, user, inventory, payment), each with independent databases and event‑driven synchronization.

9. Artificial Intelligence & AIOps : Distinguish supervised, unsupervised, and reinforcement learning. Highlight AIOps benefits: mean‑time‑to‑detect (MTTD) reduced from 45 minutes to 3 minutes and operational labor cut by 50%.

10. IoT and edge computing : Learn the three‑layer IoT stack (perception, network, application) and protocols MQTT/CoAP. Edge computing reduces bandwidth by 70% and latency from 200 ms to under 10 ms when moving video analysis to the edge.

11. Blockchain basics : Remember decentralization, immutability, transparency, and smart contracts. Differentiate public, private, and consortium chains, and consensus mechanisms (PoW, PoS, PBFT). Typical exam focus is on 1–2 multiple‑choice questions.

12. DevOps & CI/CD : Grasp the DevOps culture of development‑operations collaboration, automation pipelines, and continuous integration/delivery. Example: implementing a CI/CD pipeline shortens release cycles from two weeks to multiple daily releases and cuts production fault rate by 60%.

13. Study allocation : Mandatory content occupies ~60% of study time, core focus ~30%, and general topics ~10%. Prioritize the listed essential areas to maximize exam performance.