Katkat
Back

Design a company-wide monitoring system

LinkedInSoftware Engineer · Onsite
MediumSoftware EngineerOnsiteSystem Design

Last updated

This question evaluates skills in designing large-scale observability and monitoring systems, covering telemetry collection (metrics, logs, traces), high-throughput ingestion, storage and scaling strategies, alerting pipelines, and operational trade-offs like cardinality, retention, and backpressure.

You are asked to design a monitoring system used company-wide.

Goals

  • Collect and query telemetry for many services/hosts
  • Support alerting and dashboards for engineers and SREs
  • Handle high ingestion volume and multi-team usage

Telemetry types to cover

  • Metrics (time series)
  • Logs
  • Distributed traces (optional but strong)

Discuss

  • Key requirements (SLOs, latency, retention, tenancy)
  • High-level architecture and data flow
  • Storage choices and scaling
  • Alerting pipeline and reliability
  • Operational concerns (cardinality, cost controls, backpressure)