Search

Lessons (18)

• Observability: Logs, Metrics & Traces

  … Metrics, the "what" Metrics are numbers over time , cheap to store and perfect for dashboards and alerts. Type …

• Scaling & Health Probes

  … It also needs metrics-server installed. Scaling on other metrics: requests per second, queue length, Kafka consumer lag …

• Design a Logging and Monitoring Platform

  … agents batch into an ingestion layer, durable streams feed hot search, metrics and cold archival](/img/system-design …

• API Gateway, BFF & Service Discovery

  … networking as infrastructure As services multiply, every one needs retries, timeouts, mutual TLS, metrics and tracing. Instead of …

• Deploying & Operating Microservices

  … If metrics get worse, roll back automatically. mermaid flowchart LR U[Users] -- R{{Router}} R -- 95% V1[v1 …

• Releases, Observability and Production Safety

  … Correlate logs, metrics and traces with request IDs; alert on SLO burn, not noise. Protect traffic with TLS …

• System Design Interview Framework

  … Keep the first design simple, name what is intentionally deferred, and explain which metric says a trade-off …

• Debug Latency Like an Investigator

  … Apply the smallest reversible mitigation and verify the same metric improves. 5. Fix root cause and alert on …

• The L6 System-Design Thinking Loop

  … Explain metrics, mitigation, and the durable fix. 💡 A good trade-off is specific: “I accept a few seconds …

• Delivery Guarantees: At-Most, At-Least & Exactly Once

  … restart from offset 8 → message 7 is LOST Use for: metrics or logs where losing a little is …

• Helm, kubectl Cheat Sheet & Best Practices

  … vault NetworkPolicy scanned, pinned image Operations structured logs metrics + alerts Helm / Kustomize in git GitOps deploys - [ ] At least …

• How Browsers Render a Page

  … That's what the INP metric measures (see Frontend Performance ). Key takeaways - HTML → DOM, CSS → CSSOM, combined into …

• What Is Apache Kafka?

  … Activity tracking clicks, page views, searches Log and metrics aggregation logs from 1,000 servers Stream processing real …

• Pods

  … log shippers, service-mesh proxies (Envoy in Istio), config reloaders, metrics exporters. Init containers Run before the main …

• Consumers & Consumer Groups

  … the key health metric Lag = latest offset in the partition − committed offset of the group. It tells you …

• Scaling, Load Balancing & Rate Limiting

  … Microservices → Resilience Patterns . Back-of-the-envelope numbers Metric Rough figure --- --- 1 million requests/day ≈ 12 requests/second …

• Normalization & Denormalization

  … freshness delay. 4. Rebuild and reconciliation procedure. 5. Metric for stale/failed projections. Interview-ready answer I normalize …

• Frontend Performance & Core Web Vitals

  … their "good" thresholds](/img/frontend/web-vitals.svg) Metric Measures Good Poor --- --- --- --- LCP : Largest Contentful Paint When the …