Every European monitoring tool reviewed
Amsterdam, Netherlands
Free tier / custom pricing
Free tier
Best for: Teams that want traces, logs and infrastructure data in one store
Elastic APM's advantage is where it lives. Because it is built on the Elastic Stack, traces, metrics, logs and infrastructure data sit in the same store and correlate directly — so a slow transaction links to the exact log lines produced inside that request, instead of switching tools and reconstructing a timeline from timestamps. That single capability is what most incident investigations actually need.
The coverage is complete: distributed tracing across service boundaries, APM agents for the common languages and frameworks, error and exception tracking, real user monitoring measuring what browsers genuinely experienced rather than what synthetic checks report, and machine-learning anomaly detection over the metrics.
Elastic N.V. is based in Amsterdam, and deployment can be self-managed on your own infrastructure or on Elastic Cloud in EU regions — the deciding capability for anyone whose logs contain personal data, which is most organisations.
There is a free tier with custom pricing above it. The costs are real: the Elastic Stack is heavy to run well, tuning and capacity planning are ongoing work, and licensing has shifted over the years in ways worth checking against your intended use.
What Elastic APM does well
- Traces, logs, metrics and infrastructure in one correlated store
- Distributed tracing with agents across common languages
- Real user monitoring and ML anomaly detection
- Self-managed or Elastic Cloud in EU regions
- Dutch company, free tier available
Where Elastic APM falls short
- The Elastic Stack is heavy to run well
- Ongoing tuning and capacity planning
- Licensing has changed over the years — check your use
- Pricing above the free tier is custom
Standout feature. Click from a slow span to the log lines inside it — the one thing every incident investigation needs and most stacks cannot do.
Berlin, Germany
Free tier / from €30 per month
Free tier
Best for: Developer teams monitoring APIs and user journeys as code
Checkly's browser checks are Playwright tests, and that choice solves the problem that makes synthetic monitoring decay. Teams already write Playwright for end-to-end testing; reusing those scripts as monitoring checks means the two never drift apart, and a check exercises a genuine user journey — log in, search, add to basket, check out — rather than confirming a URL returns 200 while the journey behind it is broken.
Monitoring as code through the Checkly CLI is the structural half. Checks live in the repository, get reviewed like code and deploy through CI/CD, so monitoring changes alongside the feature it covers instead of being updated by whoever remembers. Shift-left integration means a broken journey surfaces in the pipeline rather than in production, and API monitoring handles multi-step checks with assertions across chained requests.
Alerting and incident response integrations route failures to the usual channels. Checkly GmbH is based in Berlin with EU data handling under GDPR, offering a free tier and paid plans from €30 per month. Its scope is deliberately narrow: this is synthetic and API monitoring, not infrastructure metrics or application tracing, and pricing scales with check frequency and volume, so high-frequency checks across many journeys add up.
What Checkly does well
- Browser checks written as Playwright tests, reusing E2E work
- Monitoring as code in your repository via CLI
- CI/CD integration catches breakage before production
- Multi-step API checks with assertions
- German company, EU data handling, free tier
Where Checkly falls short
- Narrow scope — no infrastructure or APM
- Pricing scales with check frequency and volume
- Requires Playwright familiarity for browser checks
- Not open source or self-hostable
Standout feature. Your end-to-end tests become your monitoring — one set of scripts that cannot drift apart from the product.
Berlin, Germany
Free (self-hosted) / from €12 per month
Free self-hosted
Best for: Teams wanting Sentry-compatible error tracking they can self-host cheaply
GlitchTip is Sentry-compatible, which is the whole practical argument: existing Sentry SDKs point at it with a configuration change rather than a re-instrumentation across every service. For a team already emitting Sentry events, migration is close to free.
What makes self-hosting realistic is the architecture. Self-hosted Sentry is notoriously heavy, needing a substantial stack of services and enough hardware that many teams give up and pay for the cloud version. GlitchTip is deliberately simplified and runs on a fraction of those resources, which turns self-hosted error tracking from a project into a container.
It covers error tracking and issue management, uptime monitoring and performance monitoring, with alerting and notification integrations. It is open source and free self-hosted, with a cloud option from €12 per month, based in Berlin. The honest trade: it is a smaller project than Sentry with a smaller community and fewer features — session replay, advanced performance analysis and the deeper integrations are not there. For most teams, error tracking is what they actually used Sentry for.
What GlitchTip does well
- Sentry-compatible — existing SDKs work with a config change
- Far lighter than self-hosted Sentry to run
- Error tracking, uptime and performance monitoring included
- Free and open source self-hosted, cloud from €12/month
- German-based with EU cloud hosting
Where GlitchTip falls short
- Fewer features than Sentry — no session replay
- Smaller project and community
- Performance monitoring lighter than a full APM
- Self-hosting still means running it
Standout feature. Sentry's SDKs pointed at something you can actually self-host without a dedicated cluster.
Berlin, Germany
Free tier / from €20 per month
Free tier
Best for: Teams standardising on OpenTelemetry who want to keep their options open
Uptrace is built on OpenTelemetry natively rather than importing from it, and that decides how expensive changing your mind later will be. Instrumentation in your code belongs to the open standard rather than to a vendor agent, so moving to a different backend becomes a configuration change instead of a re-instrumentation project across every service — which is how observability lock-in actually works in practice.
The platform covers the three signals in one place: distributed tracing across service boundaries, metrics monitoring and log management, with alerting, notifications and dashboards over all of them. Language and framework support follows OpenTelemetry's, which is broad and growing rather than dependent on one vendor writing agents.
It is open source and self-hostable, so traces and logs containing user identifiers, request payloads and internal architecture can stay on your own infrastructure, with Uptrace Cloud from €20 per month above a free tier. Uptrace is based in Berlin with EU hosting under GDPR. The realistic caveats: it is younger and smaller than Elastic with a correspondingly smaller ecosystem and community, and self-hosting means operating the storage layer yourself.
What Uptrace does well
- OpenTelemetry-native — instrumentation is vendor-neutral
- Traces, metrics and logs in one platform
- Open source and self-hostable
- Broad language support via OpenTelemetry
- German-based, cloud from €20/month above a free tier
Where Uptrace falls short
- Younger and smaller than Elastic
- Smaller ecosystem and community
- Self-hosting means operating the storage layer
- Fewer prebuilt dashboards and integrations
Standout feature. Vendor-neutral by construction: leaving costs a config change rather than re-instrumenting fifty services.
Engineering headquarters in Linz, Austria
Founded 2005
Usage-based, quoted on request
Free trial
Best for: Large estates where nobody can hold the whole architecture in their head
Dynatrace was founded in Linz in 2005 and is still engineered there, and it solves the problem the rest of this category leaves to you: knowing what your system actually consists of. Its agent discovers services, processes, containers and the dependencies between them automatically, and keeps that map current as things change.
That matters because in a large estate nobody knows the real architecture. The diagram is out of date, the person who built the integration has left, and the dependency that breaks at three in the morning is one nobody remembered existed. Automatic discovery plus AI-assisted root cause analysis is the difference between an incident where five teams check their own dashboards and one where the answer arrives with the alert.
The jurisdictional position needs stating precisely, because it is not simple: Dynatrace was founded in Austria, its engineering headquarters and research lab remain in Linz, and the company is US-listed and US-incorporated.
That is a stronger European connection than most enterprise software has and it is not the same as an EU-established vendor, so which matters is a procurement question. Data region can be chosen including the EU. Pricing is usage-based and quoted, it is the most expensive option in this category by a distance, and for a small estate it is substantial overkill.
What Dynatrace does well
- Automatic discovery of services and dependencies
- AI-assisted root cause analysis across the stack
- Founded and still engineered in Linz, Austria
- Choice of data regions including the EU
- Handles estates too large to map by hand
Where Dynatrace falls short
- US-listed and US-incorporated despite Austrian engineering
- Most expensive option in this category
- Usage-based pricing quoted rather than published
- Substantial overkill for a small estate
Standout feature. It works out what your architecture actually is — which nobody at a large organisation reliably knows.
Riga, Latvia
Founded 2005
Free and open source; paid support and services
Free to download
Best for: Large infrastructure estates where per-host pricing is the problem
Zabbix is an open-source monitoring platform from Riga that has been used at genuine scale since 2005: tens of thousands of hosts, agent-based and agentless checks, distributed proxies for remote sites, and no licence cost at any of it.
The economics are the argument. Commercial monitoring is priced per host or per ingested gigabyte, which means the estates that most need monitoring are the ones where it becomes unaffordable — and where teams start monitoring selectively. Zabbix removes that trade-off entirely: what you pay for is the hardware it runs on and, optionally, support from Zabbix SIA.
Zabbix SIA is based in Riga with a subsidiary in Tokyo, and the software is GPL. Running it yourself means telemetry about every machine you own — which is a map of your infrastructure — never leaves your network.
What Zabbix does well
- Free and open source with no host or metric limits
- Proven at very large scale
- Agent, agentless, SNMP and IPMI collection
- Distributed proxies for remote and segmented sites
- Latvian company, EU jurisdiction, commercial support available
Where Zabbix falls short
- Setup and tuning is real work
- Interface is functional rather than pleasant
- APM and tracing are not its strength
- You operate the database it depends on
Standout feature. Monitoring an estate of any size without a per-host bill.
Munich, Germany
Founded 2008
Free Raw Edition; paid editions priced per monitored service
Free trial
Best for: Teams that want monitoring configured without configuring it
Checkmk is a Munich monitoring platform whose defining feature is discovery: point it at a host and it works out which services are worth monitoring, with thresholds that are usually right, from a catalogue of more than two thousand plugins.
That matters because the failure mode of infrastructure monitoring is not a missing feature — it is an estate half-configured, where the host added last month has no checks on it. Automatic discovery makes coverage the default state rather than a maintenance task, which is a different operational reality from a system where every check is declared by hand.
Checkmk GmbH is based in Munich, grown out of Mathias Kettner GmbH, with a free open-source Raw Edition alongside paid editions priced per monitored service. Both run on your own infrastructure.
What Checkmk does well
- Automatic service discovery with sensible defaults
- Over 2,000 plugins out of the box
- Free open-source Raw Edition
- Self-hosted, so telemetry stays internal
- German company with commercial support
Where Checkmk falls short
- Paid editions priced per monitored service, which grows
- Less suited to application tracing than to infrastructure
- The GUI hides complexity you eventually meet
- Migration from another system is a project
Standout feature. Discovery that gets the estate covered without a configuration backlog.
Nuremberg, Germany
Founded 2009
Free and open source; paid subscription for support
Free to download
Best for: Teams that want their monitoring defined as code
Icinga is an open-source monitoring stack from Nuremberg, originally a fork of Nagios and now its own platform, built for teams who would rather express monitoring as configuration in version control than click it into a UI.
Config as code is the point. Hosts, services, checks and notification rules are declared in files that live in Git alongside the infrastructure they describe, which means monitoring is reviewed, diffed and deployed like everything else — and a rebuilt environment comes back monitored rather than needing someone to remember.
Icinga GmbH was founded in 2018 in Nuremberg to fund development and product management, with NETWAYS providing consulting, training and hosting around it. The software is open source and runs on your own infrastructure.
What Icinga does well
- Open source with a paid support subscription
- Monitoring defined as code, versioned in Git
- Distributed setups across zones and sites
- Compatible with the Nagios plugin ecosystem
- German company, self-hosted deployment
Where Icinga falls short
- Steeper learning curve than a discovery-led tool
- You build the dashboards you want
- Metrics and graphing rely on companion components
- Less turnkey than the commercial platforms
Standout feature. Monitoring that lives in version control with everything else.
Paris, France
Free self-hosted IT-100 edition up to 100 hosts (no time limit); self-hosted Business from €283/month (125 hosts); SaaS IT Edition from €425/month (125 hosts) and Pro from €1,125/month (250 hosts), billed annually
Free trial (Infra Monitoring and Experience Monitoring)
Best for: Infrastructure teams who want a free tier that stays free
Centreon is a Paris-built infrastructure monitoring platform with an open-source core — Apache 2.0, created by Julien Mathis and Romain Le Merlus — and a self-hosted IT-100 edition that stays free for up to 100 hosts with no time limit, which is a genuinely usable starting point rather than a crippled trial.
Beyond that free tier, the commercial editions add SLA monitoring, geographical and business views, and reporting: self-hosted Business starts from €283 per month for 125 hosts, and a SaaS IT Edition from €425 per month for the same footprint, so the same product scales from a self-hosted free tier to a managed cloud deployment without switching tools. More than 700 monitoring connectors cover the usual mix of cloud, network and application targets out of the box.
Centreon SAS is based in Paris. The trade-off against Zabbix or Icinga is that the free tier caps out at 100 hosts before the pricing kicks in, and against Checkmk the discovery experience is less automatic — Centreon expects more deliberate configuration of what to monitor.
What Centreon does well
- Open-source core (Apache 2.0), free self-hosted up to 100 hosts
- 700+ monitoring connectors out of the box
- Same product as self-hosted or SaaS
- French company, EU jurisdiction
- SLA monitoring and business views on paid editions
Where Centreon falls short
- Free tier capped at 100 hosts
- Less automatic discovery than Checkmk
- SaaS hosting region not published
- Commercial pricing scales with host count
Standout feature. A free self-hosted tier that actually stays free up to 100 hosts, not a time-limited trial.
Nuremberg, Germany
PRTG 500 from $200/month, PRTG 1000 from $358/month, PRTG 2500 from $742/month, PRTG 5000 from $1,300/month, PRTG 10000 from $1,642/month, billed annually; also available as PRTG Hosted Monitor or on-prem Enterprise Monitor for larger environments
30-day fully-featured free trial, no credit card
Best for: Network teams who want one sensor-based tool for everything
PRTG, from Paessler in Nuremberg, monitors servers, networks, cloud services and OT from one sensor-based console, and its age shows up as maturity rather than dust: the sensor library covers most common infrastructure out of the box, and a 30-day fully-featured trial with no credit card lets a team try the real product before licensing anything.
Licensing is sensor-count based rather than per-host, which is a different shape from Centreon or Checkmk: PRTG 500 starts at $200 per month for up to roughly 50 devices, scaling to PRTG 10000 at $1,642 per month for around 1,000 devices, billed annually. Deployment is either self-hosted or as PRTG Hosted Monitor, a managed option, with PRTG Enterprise Monitor for estates beyond 1,000 devices.
Paessler GmbH is registered in Nuremberg. Against Zabbix or Icinga, PRTG is not open source and the per-sensor pricing model can get expensive as an estate grows; against Checkmk, its discovery is more manual, sensor by sensor.
What Paessler PRTG does well
- One console for servers, network, cloud and OT
- 30-day fully-featured trial, no credit card
- Self-hosted or PRTG Hosted Monitor
- Mature sensor library covering common infrastructure
- German company, EU jurisdiction
Where Paessler PRTG falls short
- Not open source
- Sensor-based pricing can grow expensive at scale
- US-dollar pricing on a German product
- Less automatic discovery than Checkmk
Standout feature. One sensor-based console for servers, network, cloud and OT, with a trial that is the real product.
Berlaar, Belgium
Founded 2018
Solo $17/month (2 sites/servers) up to Scale $439/month (200); Enterprise on request; two months free when billed annually
10-day free trial, no credit card
Best for: Small teams and agencies who want uptime, SSL and broken links in one check
Oh Dear, built by the Spatie developers Freek Van der Herten and Mattias Geniar in Belgium, treats a website as more than an uptime number: uptime checks from multiple locations sit alongside SSL certificate expiry, broken-link crawling, mixed-content detection, DNS and domain-expiry watching, and scheduled-task tracking, so one monitor replaces several single-purpose tools for a small team or agency running client sites.
Monitors can be defined in plain English with AI verification, and the product leans API-first for teams that want monitoring as code rather than a manual dashboard. Primary data is kept in Belgium with Combell, an ISO 27001-certified host, with performance metrics stored separately via ClickHouse Cloud.
Immutable VOF operates from Berlaar, Belgium. Pricing runs from $17 per month for 2 sites or servers up to $439 per month for 200, with a 10-day free trial. Against Checkly or Better Stack, Oh Dear stays narrower — it covers website and endpoint health rather than distributed tracing or infrastructure metrics, and there is no self-hosted option.
What Oh Dear does well
- Uptime, SSL, broken links and DNS in one check
- AI-verified, plain-English monitor definitions
- Primary data hosted in Belgium (ISO 27001 host)
- API-first, monitoring-as-code friendly
- Belgian company, EU jurisdiction
Where Oh Dear falls short
- No self-hosted option
- Narrower scope than a full APM or infra platform
- Pricing scales with number of sites/servers
- Performance metrics stored via a separate subprocessor
Standout feature. One check that covers uptime, SSL, broken links and DNS — built by developers for their own client sites.
Grays, England, United Kingdom
Founded 2012
Free plan with 10 monitors; Superior from €16.66/month billed annually (€19.99 monthly); Business from €58.33/month billed annually (€69.99 monthly); Enterprise on request
7-day free trial, no credit card
Best for: Teams that want a straightforward, free-tier uptime monitor
StatusCake, run by TrafficCake Limited since 2012, is a website and uptime monitor built around a genuinely usable free plan: 10 uptime monitors at 5-minute intervals, plus a page speed, domain and SSL monitor, before any payment is required — a reasonable way to start monitoring a small site without committing to a vendor.
Paid plans tighten the check interval and widen coverage: Superior, from €16.66 per month billed annually, drops to 1-minute checks across 100 monitors with domain and SSL monitoring added; Business, from €58.33 per month, goes to 30-second checks across 300 monitors with server monitoring and team tools. Checks run from AWS, Google Cloud and DigitalOcean infrastructure.
TrafficCake Limited is registered in Grays, England — UK jurisdiction, outside the EEA under an adequacy decision. The data processing agreement notes personal data may be processed outside the UK/EEA, which is worth checking against your own requirements. Against Oh Dear or Better Stack, StatusCake stays purely in uptime and website monitoring rather than broader observability.
What StatusCake does well
- Usable free plan with 10 monitors
- Check intervals down to 30 seconds on paid plans
- Page speed, domain and SSL monitoring included
- UK company, adequacy-decision jurisdiction
- 7-day free trial on paid plans
Where StatusCake falls short
- Data may process outside the UK/EEA
- No self-hosted option
- Narrower scope than a full observability platform
- No published security certifications
Standout feature. A free plan with 10 real monitors, not a crippled trial, before any payment is required.
