// category

Software Architecture

How real systems are designed — patterns, trade-offs, scaling, and the decisions that hold up under load. Architecture for engineers who have to ship and maintain what they draw.

23 guides~1.1 hrs totalupdated Jul 6, 2026

 Start here

CDNs and Object Storage: Store at the Origin, Serve at the Edge (2026)

CDNs and object storage explained in 2026 — object storage as the origin for large files, CDNs as the edge cache near users, how they work together, and what they store in AI systems.

Beginner · 5 min

Software Architecture

Read the guide →

All guides

12 on this page

BeginnerRate Limiting Explained: Token Bucket, Leaky Bucket, and When Each Wins (2026)Rate limiting explained in 2026 — fixed window, sliding window, token bucket, and leaky bucket, which algorithm to default to, and why LLM traffic is limited by tokens rather than requests.5 min BeginnerHorizontal vs Vertical Scaling: Scale Up Until You Can't (2026)Horizontal vs vertical scaling in 2026 — scale up (a bigger box) versus scale out (more boxes), the hardware ceiling, why horizontal needs statelessness, and the diagonal middle ground.5 min BeginnerCAP Theorem Explained: It Was Never 'Pick Two' (2026)CAP theorem explained in 2026 — why 'pick two of three' is wrong, the real consistency-vs-availability choice during a partition, PACELC, and what it means for AI systems.5 min BeginnerEvent-Driven Architecture: Trading Control for Decoupling (2026)Event-driven architecture explained in 2026 — events vs commands, choreography vs orchestration, and the real trade-off: you gain decoupling and lose the single view of what happens next.5 min BeginnerAPI Gateway Explained: It's Not a Load Balancer (2026)API gateway explained in 2026 — what it actually does (auth, rate limiting, routing, aggregation), why it's not a load balancer, and where the new AI gateway fits.4 min BeginnerSQL vs NoSQL: It's About Access Patterns, Not Scale (2026)SQL vs NoSQL in 2026 — why the real choice is about access patterns and correctness, not scale; the four NoSQL types; and why most systems end up using both.5 min BeginnerLoad Balancers Explained: Health Checks Do the Real Work (2026)Load balancers explained in 2026 — L4 vs L7, the algorithms, and why health checks (not even distribution) are the part that actually keeps you online.5 min BeginnerSystem Design Building Blocks: A Decision Map, Not a Parts List (2026)System design building blocks explained as a decision map — the five questions that pick your components (architecture style, cache, message broker, database, load balancer) and how each block earns its place.6 min BeginnerCaching Strategies: Why Invalidation Is the Hard Part (2026)Caching strategies in 2026 — the patterns (cache-aside, write-through, write-behind), eviction policies, and why cache invalidation, not adding a cache, is the genuinely hard part.7 min BeginnerKafka vs RabbitMQ: A Log vs a Queue, Not Fast vs Slow (2026)Kafka vs RabbitMQ in 2026 — why it's a log versus a queue (not fast versus slow), the differences that matter, and when each one (or both) is the right call.7 min BeginnerMonolith vs Microservices: It's About Team Size, Not Traffic (2026)Monolith vs microservices in 2026 — the differences that matter, why team size (not traffic) decides it, and when to start with a modular monolith.7 min

Page 1 of 2Older →

Following Software Architecture?

Get new Software Architecture guides the day they publish. One email, no spam.