Self-Hosted Datadog Alternative for Docker and Kubernetes

The Datadog Problem

Datadog is a fantastic product. It is also one of the most expensive monitoring solutions on the market.

Datadog bills per host, not per container, and the list prices are public: Infrastructure Pro is $15 per host per month billed annually ($18 on demand), Enterprise $23, and APM adds $31 per host per month on top. Logs, custom metrics and synthetics are metered separately.

The number that surprises teams is not the unit price, it is the multiplication. Ten hosts on Infrastructure Pro plus APM is roughly $5,500 a year before a single log line is ingested; the same fleet with log management and custom metrics routinely lands several times higher. Because billing follows the high-water mark of your hourly host count, autoscaling pushes the bill up faster than average utilisation suggests.

And beyond cost, there is a fundamental trade-off: all your infrastructure data leaves your network. Every container state, every log line, every metric data point flows to Datadog’s cloud. For many organizations (especially in healthcare, finance, or government) that is a non-starter.

Who Should NOT Switch

Let’s be upfront. Maintenant does not replace Datadog if you need:

• APM / distributed tracing: Datadog’s APM is a full OpenTelemetry-compatible tracing platform. Maintenant does not do application-level tracing.
• Log aggregation at scale: Datadog can ingest terabytes of logs with full-text search. Maintenant shows container logs in real-time but does not store or index them.
• 700+ integrations: Datadog integrates with AWS, GCP, Azure, databases, message queues, and everything in between. Maintenant monitors the Docker/Kubernetes layer.
• Multi-region, multi-cloud observability: Datadog is designed for distributed infrastructure across regions. Maintenant is per-host or per-cluster.

If you are a large engineering organization running thousands of microservices across multiple cloud providers, Datadog (or a comparable enterprise tool) is likely the right choice.

Who Should Switch

For a surprising number of teams, Datadog is overkill:

• Self-hosters running 5-100 containers on 1-3 VPS instances
• Small teams who need “are my services up?” not “what is the p99 latency of my gRPC calls?”
• Homelab enthusiasts who want monitoring without a per-host SaaS subscription
• Privacy-conscious organizations that cannot or will not send infrastructure data to a third party
• Startups burning runway on monitoring bills that exceed their hosting costs

If you recognize yourself here, read on.

Maintenant vs Datadog: Feature Comparison

The Numbers

What You Get with Maintenant

One container. Zero configuration files. Everything monitored:

• Container auto-discovery: Docker and Kubernetes, states, health checks, restart loops, Compose projects grouped automatically

• HTTP/TCP endpoint monitoring: configured via Docker labels, no YAML
• SSL/TLS certificate tracking: automatic detection, expiration alerts at 30/14/7/3/1 day

• Heartbeat / cron job monitoring: unique ping URL per job, duration tracking, missed execution alerts (5 free, unlimited from Personal)
• System resources: CPU, RAM, network, disk per container and per host

• Update intelligence: OCI digest scan, detect available updates across all containers
• Unified alerts: webhook, Discord (Community), Email (Personal), Slack and Teams (Pro)
• Public status page: real-time, SSE updates, incident management (Personal)

• Network security insights: open ports, exposed services, unencrypted endpoints (Pro: unified security posture with CVE enrichment and risk scoring)
• REST API + MCP server: full programmatic access, AI-assistant ready

Migrating from Datadog

There is no migration per se: Maintenant auto-discovers your containers the moment you deploy it. The switch takes 30 seconds:

# docker-compose.yml: add this service
services:
  maintenant:
    image: ghcr.io/kolapsis/maintenant:latest
    ports:
      - "8080:8080"
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock:ro
      - /proc:/host/proc:ro
      - maintenant-data:/data
    environment:
      MAINTENANT_ADDR: "0.0.0.0:8080"
      MAINTENANT_DB: "/data/maintenant.db"
    restart: unless-stopped

volumes:
  maintenant-data:

docker compose up -d

Open http://your-server:8080: your containers are already there. Add Docker labels to monitor HTTP endpoints:

labels:
  maintenant.endpoint.url: "https://api.example.com/health"
  maintenant.endpoint.interval: "60s"

Once you confirm Maintenant covers your needs, remove the Datadog agent and stop paying.

FAQ

Can I run Maintenant alongside Datadog during a transition? Yes. Maintenant is read-only: it mounts the Docker socket in read-only mode and does not interfere with Datadog’s agent or any other tool.

Does Maintenant support Kubernetes? Yes. Maintenant auto-detects whether it runs on Docker or Kubernetes and adapts accordingly.

What about long-term metric retention? History is served as fixed windows, not as an arbitrary query range. Community opens 1h · 6h · 24h · 7d. Personal opens 1h · 6h · 24h · 7d · 30d. Pro opens 1h · 6h · 24h · 7d · 30d · 90d. The points coarsen as the window widens — Community: Raw, then 1 min; Personal: 5 min, then 1 h; Pro: 1 day. If you need longer, export via the REST API to your storage of choice.

Is there a free tier? Yes. Maintenant Community is free forever, open source (AGPL-3.0), and includes all core monitoring: containers, endpoints, heartbeats, certificates, the public status page, and the full Swarm and Kubernetes views. Personal lifts the caps and adds email alerts, CVE enrichment, security posture and incidents for 149 € once. Pro adds Slack, Teams, escalation and the right to commercial use, at 29 €/month.

Where is my data stored? On your server. Maintenant uses an embedded SQLite database. Nothing leaves your infrastructure. Ever.

Related reading

• Docker monitoring: the complete guide
• Prometheus + Grafana, the other self-hosted route
• self-hosted monitoring in 2026