← all projects
Go
PostgreSQL
Redis
Next.js
Docker
Grafana
Live2025· Backend Engineer
Uptime Monitoring Platform
Distributed uptime, latency, and incident monitoring.
GoPostgreSQLRedisNext.jsDockerGrafana
Pulse runs checks from multiple regions, stores time-series latency in TimescaleDB/Postgres, and turns failures into smart, deduplicated alerts with a public status page.
3+
Check regions
30s
Resolution
Yes
Alert dedup
// case study
Pulse — Uptime Monitoring Platform
Pulse runs HTTP/TCP checks from multiple regions, records latency as time-series data, and turns failures into deduplicated, escalating alerts — fronted by a public status page.
Architecture
Lightweight Go probes run on a schedule in each region and publish results to a central collector. Results land in Postgres (TimescaleDB hypertables) for fast windowed queries; Redis backs the dedup + escalation state.
A check definition is intentionally simple:
yaml
checks:
- name: api
url: https://api.example.com/health
interval: 30s
regions: [iad, fra, sin]
expect:
status: 200
latency_ms_p95: 400Dashboard
Features
- Multi-region checks with per-region latency breakdowns.
- Smart alerts — failures are deduplicated and escalated, not spammed.
- SLA + p95 rollups over arbitrary windows.
- Public status page generated from the same data.
Links
// links
