Self-Hosted Monitoring
One server that keeps an eye on all of yours, and it stays under your control. You learn about problems before your users do, not after something has already gone down. The monitoring data stays on your own server. Set up on a seed in minutes.
Why self-hosted monitoring?
Monitor your own seeds and servers
Keep CPU, memory, disks, processes and services across your whole fleet in view. Whether a single seed or dozens of servers, everything comes together in one interface.
Monitor external services and endpoints
Watch things that do not run at dataforest too: websites, APIs, mail servers or the availability of other providers, checked from a location outside your own infrastructure.
Early warning instead of outage
Full disks, soon-to-expire TLS certificates, stopped services or unusual load: your monitoring flags it before any of it becomes a real outage. You fix problems calmly instead of chasing them under pressure.
One dashboard for everything
Instead of scattered status pages and gut feeling, you get one overview that shows at a glance what is green and what needs attention. A single place for the health of your entire infrastructure.
The case for self-hosted monitoring
Your data, your server
All metrics stay on your server and under your control. You decide what is monitored, how long data is kept and who sees it. That also keeps the GDPR assessment simple: no third party processes telemetry about your infrastructure.
Flat, predictable cost
Whether you monitor 3 or 30 hosts: the price stays that of your seed. No billing per host, per metric or per gigabyte of logs, the invoice does not grow with your environment.
No vendor lock-in
Open source, your configuration, standard protocols. You are tied to no proprietary format and no vendor, and you move your monitoring elsewhere at any time.
Full data history
How long metrics are kept is decided solely by your disk. The full history stays available for capacity planning and trend analysis.
One server, many hosts
A single monitoring server watches anywhere from a few to several hundred hosts. When new infrastructure shows up, you add it to the same server instead of standing up a second system.
More than just monitoring
The seed is fully yours: backups, cron jobs or internal tools run on the same server alongside the monitoring. You pay for one server, not one service per task.
An outage that never happened
Without monitoring you would have noticed only once the service had already crashed. With monitoring the problem had long been fixed before anyone noticed.
02:14 Disk at 85 percent
cache-01 reaches the warning threshold
02:15 Trend detected
At this rate the disk will be full in about 40 minutes
02:16 Notification sent
Email and chat message to the team
02:21 Logs rotated
You free up space before it turns critical
02:25 Status green again
No outage, no data loss, nobody noticed a thing
Self-hosted monitoring or a monitoring SaaS?
Self-hosted monitoring
Best choice when data sovereignty, predictable cost and independence matter.
- Full data sovereignty: telemetry about your infrastructure stays on your server and under your control
- Flat, predictable cost: you pay only for the seed, no billing per host, per metric or per gigabyte of logs
- No vendor lock-in: open source, your config, standard protocols, portable at any time
- Retention on your terms: disk-bound and effectively unlimited instead of capped by plan tier
- Monitor private and internal targets directly, without exposing them to the outside
- You run, patch and back up the server yourself and own its uptime, monitoring the monitor
- Some initial setup effort, and scaling is up to you
Monitoring SaaS
A good fit when you do not want to run a server and need to start immediately.
- Zero maintenance: nothing to patch, back up or keep highly available, the vendor handles it
- Ready in minutes with no infrastructure of your own, a polished UI and a large catalog of integrations
- Managed high availability and scaling, the vendor guarantees the monitoring system's own uptime
- Support and an SLA on paid tiers
- Pricing scales with hosts, custom metrics, log volume and add-on features, which makes it harder to predict
- Telemetry about your infrastructure sits in the vendor's cloud, the GDPR assessment depends on vendor and region
- Retention is limited depending on the plan, and dashboards, alerts and agents live in the vendor's format
Self-hosted monitoring suits anyone who wants full control over their monitoring data and predictable cost, and is willing to run a server in return. A monitoring SaaS is the better choice when convenience and instant availability come first.
The ideal place for your monitoring server
We operate our own network (AS58212) with redundant infrastructure. For your monitoring server that means low latency to external targets, high availability and a location that is independent of the infrastructure you watch.
Independent Location
Open Peering Policy
Own Autonomous System
99.98%+ Availability
Checkmk, Zabbix or Prometheus + Grafana?
Strengths
- Extremely mature: publicly developed since 2001, large community, LTS releases supported until 2029
- No feature gating: every feature ships in the free edition, the vendor earns through support rather than licenses
- Broad coverage from one product: agent, SNMP, IPMI, JMX and HTTP checks with hundreds of official templates
Weaknesses
- Needs a full stack on the server: server daemon, PHP frontend and a separate MySQL or PostgreSQL database
- Steep learning curve: the model of items, triggers, templates and macros takes real time to learn
- More manual work at the start: hosts and templates are linked by hand, service discovery is less automatic
Strengths
- Automatic service discovery: install the agent and Checkmk detects file systems, processes and services on its own
- Complete system in one package: web UI, alerting, agents and over 2,000 maintained check plug-ins out of the box
- Fastest path to a running monitoring server: install one package, create the site with a single command, done
Weaknesses
- The Community edition runs on the classic Nagios core and scales less far than the commercial editions
- Positioned by the vendor for environments up to roughly 100 hosts, beyond that the commercial Pro edition takes over
- Convenience features such as automatic agent updates and reporting are reserved for the paid editions
Strengths
- De-facto standard for metrics in cloud-native and Kubernetes environments, ready-made exporters for almost every service
- Grafana dashboards are the reference for visualization, thousands of community dashboards can be imported directly
- PromQL enables powerful queries and trend-based alerts far beyond static thresholds
Weaknesses
- A kit rather than a product: Prometheus, Grafana, Alertmanager and one exporter per service, each maintained separately
- No automatic discovery of classic services and no ready-made OK/WARN/CRIT status model, you define thresholds yourself
- Even simple alerts such as a disk filling up require PromQL knowledge and configuration across several components
All three tools have earned their place: Zabbix as the equally complete but steeper alternative, Prometheus + Grafana with its strengths in cloud-native environments. We focus on Checkmk because it is the easiest to set up: automatic service discovery, a complete system in one package, and our guides walk you through the setup step by step.
What your monitoring server looks like
Checkmk runs as its own site on the seed. One command shows the status of all services, another discovers the services of a new host and activates them.
Your own monitoring in 3 steps
Create a seed
Pick a model in the configurator that fits your environment. A small model is enough to start; it is mainly RAM that grows with your environment.
Install Checkmk
Install the Checkmk Community package through the package manager, create a monitoring site with a single command and open the web UI. Our guide walks through the setup step by step.
Add hosts and services
Roll out the agent on your first servers, let Checkmk discover the services automatically and activate the changes. A few minutes later your first services are green, part two of the guide shows every step.
Step-by-step setup
Tested on a current dataforest Seed. Every command explained, accessible without an account.
Add your first hosts and services in Checkmk
Step-by-step guide to install the Checkmk agent on a server and register it with TLS, discover services automatically, and set up notifications via email and Pushover.
Open guideSet up a Checkmk monitoring server
Step-by-step guide to install Checkmk Community on a dataforest seed, create a monitoring site and serve the web interface through Caddy with automatic HTTPS.
Open guideConfigure a seed
Billed hourly, no minimum term, no setup fee. The server is available for other services alongside monitoring.
Entry
Beginner
Standard
All-rounder
Performance
CPU-optimized
All prices incl. 19% VAT
Why dataforest cloud?
Data sovereignty
Your data stays in Germany. All seeds run in certified data centers in Frankfurt. No data transfers to third countries, full GDPR compliance.
Deployed in seconds
Seeds are provisioned automatically. From configuration to a running server takes only seconds. No waiting, no tickets.
Hourly billing
You only pay for what you use. No minimum terms, no setup fees. Seeds can be created and deleted at any time.
Full control
Root access, public API and full transparency. You decide what runs on your seed. No vendor lock-in, no hidden restrictions.
Bevor Sie loslegen.
Does the monitoring data really stay with me?
Who takes care of updates and server security?
Who monitors the monitoring server itself?
Can I monitor services that do not run at dataforest?
When does a monitoring SaaS make more sense?
What can I monitor?
Do I need an agent on every server?
How do I get notified?
How many hosts can one server handle and do I need powerful hardware?
Do I need prior knowledge?
Is the Checkmk Community edition enough?
What about Prometheus and Grafana?
Any questions?
Our experts are happy to help. You'll be surprised how fast we are.
