Search

Lessons (24)

• Capacity, Availability, Latency and CAP

  … a design chooses latency targets, availability targets and a partition-time consistency policy](/img/system-design-l6/capacity …

• Debug Latency Like an Investigator

  High CPU is not required for high latency. Follow percentile shape and trace wait time before scaling blindly …

• Observability: Logs, Metrics & Traces

  … request duration seconds What are p50, p95, p99 latencies? js import client from "prom-client"; const httpDuration = new …

• Design a Ride-Sharing Service

  … Watch stale-location age, match latency, conflicts and regional imbalance.

• Design a Logging and Monitoring Platform

  … Measure ingestion lag, dropped samples, query latency, cardinality and storage cost.

• System Design Interview Framework

  … framework.svg) 1. Clarify users, core flows, scope, latency, availability and consistency. 2. Estimate peak RPS, storage, bandwidth …

• Real-Time, Pub/Sub, Webhooks and CDC

  … retry and make receiver idempotent WebRTC Direct low-latency media/data Needs signaling, NAT traversal and fallback Pub …

• Service Communication: Sync vs Async

  … Total latency = sum of every hop. Any failure = whole request fails. ✅ Simple to understand and debug; you get …

• Caching, Consistent Hashing and Sharding

  … distribution.svg) Tool Solves Risk --- --- --- Cache Repeated read latency/load Stale data and stampedes Consistent hash ring Placement …

• Networking, DNS and Traffic Routing

  … Monitor DNS TTL/propagation, edge health, connection errors, regional latency and uneven load. A DNS answer is not …

• Deploying & Operating Microservices

  … to the new version, watching error rates and latency at each step. If metrics get worse, roll back …

• Timeouts, Retries & Backpressure

  … jobs to a visible DLQ. Watch: queue age, saturation, rejected work, dependency latency, retries, and error-budget burn.

• N+1 Queries, Loading & Pagination

  … acquisition time, returned rows and p95/p99 endpoint latency. In development, log duplicate query fingerprints. In production, sample …

• Serverless Architecture

  … runtimes (Node, Go, Rust), use provisioned concurrency for latency-critical paths, or edge runtimes (Cloudflare Workers, Vercel Edge …

• Resilience Patterns

  … AbortSignal.timeout(2000), // give up after 2 s }); 💡 Base timeouts on the dependency's p99 latency plus a …

• Caching & Redis

  … the closer the data, the faster it is](/img/backend/latency-ladder.svg) Where data lives Typical latency …

• Monolith vs Microservices

  … a system: 1. The network is reliable. 2. Latency is zero. 3. Bandwidth is infinite. 4. The network …

• Kafka vs RabbitMQ & the Kafka Ecosystem

  … topic Stream processing ❌ ✅ Kafka Streams, Flink, ksqlDB Typical latency Very low for individual messages Low, optimised for throughput …

• Backend Languages: Node.js, Go, Python & Rust

  … guaranteed by the compiler Performance-critical systems, low latency, low memory Steep learning curve, slower compile times The …

• Message Queues & RabbitMQ

  … in Node.js NATS Lightweight pub/sub Low-latency messaging Key takeaways - Queues decouple producers and consumers, smooth …

• Kafka Producers

  … batch Higher = bigger batches, better throughput, slightly more latency batch.size Max bytes per batch per partition Bigger …

• API Gateway, BFF & Service Discovery

  … every call Trade-off: more moving parts, extra latency per hop and extra resources. Worth it with many …

• Scaling, Load Balancing & Rate Limiting

  … at a fixed rate Very smooth output, adds latency In a multi-server setup, keep counters in Redis …

• Choosing an Architecture

  … automated checks for architecture rules (dependency rules, latency budgets) Guard the architecture continuously in CI ADRs for every …