How I cut my Datadog bill by 90% - the open-source observability stack I switched to

说明与用途

作者将 Datadog 换成 Prometheus、Grafana、Loki、SigNoz、Uptime Kuma 等开源监控栈，将每月费用从 6,026 美元降至 31 美元，同时保持基础设施、日志和 APM 的可见性。

问题

High cost of Datadog for 40 hosts, especially metric pricing, leading to ~$6,000/month bill.

结果

Reduced monthly cost from $6,026 to $31, a $5,995/month saving (~$72,000/year).

工作流

• Set up Prometheus + Grafana for infra metrics (Week 1).
• Add Loki + Promtail for logs (Week 2).
• Migrate alerts to Alertmanager (Week 3).
• Add SigNoz for APM on key services (Week 4).
• Decommission Datadog (Week 5).

证据

I was paying $1,800/month to Datadog for 40 hosts. Now I pay $0.

来源

Total: ~$6,000/month

来源

Setup: ~3 hours. Single VM, 2 GB RAM handles my 40 hosts comfortably.

来源

Setup: ~4 hours including Promtail config for journald, docker logs, and a few specific app files.

来源

Setup: ~2 hours (the OpenTelemetry SDK changes were the most painful part).

来源

What I gave up: Auto-instrumentation for some libraries; Real-time log streaming; Some managed integrations.

来源

What I gained: No per-metric pricing; Predictable cost; My data stays on my servers; Custom queries.

来源

About 30 hours total over 5 weeks.

来源