Design a Metrics & Monitoring System

Build an observability platform similar to Datadog or Prometheus that collects, stores, and visualizes system metrics.

Functional Requirements

• Collect metrics from thousands of services
• Store time-series data efficiently
• Query metrics with aggregations and filters
• Set up alerts based on metric thresholds
• Create dashboards for visualization
• Support for custom metrics and tags

Non-Functional Requirements

• Ingest millions of data points per second
• Query historical data spanning months
• Alert latency under 30 seconds
• Efficient storage with compression

Questions to Consider

• What storage engine is optimized for time-series data?
• How do you handle high-cardinality metrics?
• What query language will you support?