Software Architecture & System Design

By the end, you can

• Explain why "architecture" refers specifically to the decisions that are expensive to reverse.
• Contrast scaling up versus scaling out, and explain why stateless app servers are required for horizontal scaling.
• Trace a request through the standard pipeline (CDN, load balancer, app servers, cache, database, queue) and state each component's role.
• Explain the cache-aside read pattern and identify why cache invalidation is the hard part.
• Distinguish replication from sharding, and state what each technique scales.
• Compare monolith, microservices, and event-driven architectures by their wins and costs, and choose appropriately for a given context.
• Perform back-of-envelope estimates (average QPS, storage per month) given a traffic number.
• Convert an availability percentage to downtime per year, and name the mechanisms that achieve high availability.
• Apply the CAP theorem to explain the trade-off every distributed system must make.