15 Must-Know Elements of System Design

The post groups system design into a set of building blocks used in distributed systems. It covers how these pieces support scalability, fault tolerance, performance, communication between services, data storage, and visibility into system health.

Reading notes

#

• Distributed systems split a system into services that run across different servers or regions to improve scalability, fault tolerance, and performance.
• Message queues store and forward messages for asynchronous communication between services, which decouples microservices and supports independent scaling.
• Caching keeps frequently accessed data in memory so applications can return data faster, reduce latency, and lower database load.
• Task schedulers coordinate batch jobs and background tasks across distributed environments so critical jobs run reliably.
• Scalability lets a system handle higher demand, with vertical scaling adding more resources to existing servers and horizontal scaling adding more servers.
• CDNs serve cached content from geographically distributed servers near users to reduce latency and bandwidth costs.
• Consistent hashing spreads data across nodes while reducing the amount of remapping needed when nodes join or leave.
• Service discovery lets microservices find and communicate with each other dynamically instead of relying on hard-coded IPs.
• DNS turns human-readable domain names into IP addresses for web applications.
• Load balancers split incoming traffic across backend servers using methods such as round-robin, least connections, or weighted distribution.
• API gateways act as a single entry point for microservices and handle routing, authentication, and rate limiting.
• Databases store structured or semi-structured data, with SQL databases emphasizing ACID compliance and complex queries and NoSQL databases emphasizing horizontal scaling and flexible schema.
• Object storage is used for unstructured data such as images, videos, and documents, with durability and availability as key properties.
• Sharding divides a large database into smaller distributed shards based on a key to improve horizontal scalability.
• Replication copies a database to improve availability and fault tolerance, using leader-follower or multi-leader setups.
• Observability relies on monitoring, logs, and traces to make system health visible and help diagnose bottlenecks and failures.