Buyer’s Guide - August 2026

The best NGINX monitoring tools, ranked

Q: How do I enable NGINX stub_status for monitoring?

Add a location block to nginx.conf: location /stub_status { stub_status; }, restricted with allow/deny directives so only localhost or your monitoring agent’s IP can reach it. Reload with nginx -s reload or systemctl reload nginx, then verify with curl http://localhost/stub_status. Netdata auto-detects the endpoint with zero configuration; other tools require you to point them at the status URL explicitly.

NGINX is your front door: web server, reverse proxy, load balancer, and cache in one process. Monitoring it well means more than counting requests. We ranked ten tools on NGINX metric depth, collection granularity, setup effort, and how predictable the bill is when traffic grows.

Start with Netdata (free) Jump to rankings

Why this list exists

NGINX monitoring fails in two predictable ways. The first is shallow coverage: a tool reads the four stub_status counters (active connections, accepted/handled, requests) and calls it done, while upstream health, SSL handshakes, cache hit ratios, and server zones go unwatched. The second is coarse sampling: connection spikes and request microbursts that last a few seconds are invisible to a collector polling every 15 or 60 seconds, which is the default for most tools on this list.

Three dimensions decide the outcome more than any feature checklist:

Metric depth. Does the tool go beyond stub_status into the NGINX Plus API (zones, upstreams, cache, SSL, rate limiting) and into access and error logs? Logs explain what counters cannot: which requests failed and which upstream was slow.
Collection granularity. Per-second collection catches the short-lived events, like connection backlog saturation or TLS CPU spikes, that minute-averaged dashboards smooth into nothing.
Cost shape. Per-node pricing is forecastable. Per-GB and per-host models grow with traffic in ways that punish exactly the high-volume NGINX fleets this guide is for.

One more note on honesty: we do not quote list prices for any vendor except Netdata. Pricing pages change, tiers hide limits, and a dollar figure copied here would be stale within a quarter. Instead, each card describes the pricing shape and what makes the bill grow, with a link to the vendor’s current pricing page. For Netdata’s own pricing, see our pricing page. For hands-on setup material beyond this comparison, our NGINX operator guides walk through enabling stub_status, securing the endpoint, and configuring collectors.

We assembled the shortlist from tools with documented NGINX integrations: vendor documentation, the official NGINX exporter ecosystem, and independent reviews. Every tool here can collect at least the stub_status basics; the ranking separates them on what happens after that.

Metric depth and collection granularity carry the most weight (45% combined) because they determine whether you can actually diagnose an NGINX incident or just observe that one happened. Setup effort and alerting quality follow, because a tool that takes a week to configure or floods you with static-threshold noise gets abandoned. Cost predictability and deployment flexibility round out the score.

Tester credit

Compiled by the Netdata team - Updated August 12, 2026

Sources verified

Scoring criteria

NGINX metric depth and coverage 25%
stub_status basics plus Plus API metrics: upstreams, cache, SSL, zones, rate limiting
Collection granularity 20%
Per-second vs 15s/1m polling; the sharpest differentiator in this category
Setup and time to value 15%
Auto-discovery and zero-config collectors vs manual exporter and template assembly
Alerting and anomaly detection 15%
Pre-built NGINX alert rules and ML anomaly detection vs hand-written thresholds
Log correlation 10%
Access and error logs correlated with metrics to shorten MTTR
Cost predictability 10%
Per-node vs per-GB vs per-host; what makes the bill grow on high-traffic fleets
Deployment flexibility 5%
Self-hosted and air-gapped options vs SaaS-only

Vendor 01 / 10 · #netdata

Netdata

Open-source, per-second infrastructure monitoring with a zero-configuration NGINX collector and ML anomaly detection on every metric.

Best for

Teams that want NGINX metrics, logs, and host infrastructure visibility in one tool without assembling a Prometheus + Grafana stack
Buyers who need per-second granularity to catch connection spikes and request microbursts
Organizations that want predictable per-node pricing instead of usage-based bills that grow with traffic

Pricing

Per-node: Netdata Cloud Business from $4.50/node/month on annual plans, with the per-node price decreasing as node count grows
Agents are open source (AGPL) and free; a free Cloud tier covers small, non-commercial fleets
Unlimited metrics, logs, users, and retention - no per-GB, per-metric, or per-cardinality charges

Pros

Zero-configuration NGINX collector auto-detects stub_status on localhost port 80 (also /basic_status, /nginx_status, /status) with a 1-second default collection interval
Separate collectors for the NGINX Plus API, NGINX VTS, and NGINX Unit, plus web server log collection
800+ integrations cover Kubernetes, containers, and the full host alongside NGINX, so a slow upstream shows up next to the CPU or disk that caused it
400+ pre-configured alerts and ML anomaly detection on every metric, every second, with a verified 99% false-positive reduction
Zero-pipeline log management via systemd-journal and Windows Event Logs, correlated with metrics in the same dashboard
Verified customer outcomes: 80% MTTR reduction and 90% cost reduction

Where teams pair it

The basic stub_status collector exposes four core charts (connections, connection status, accepted/handled, requests); NGINX Plus metrics like zones, upstreams, cache, and SSL require enabling the separate Plus collector
No default alerts ship for the basic NGINX integration, so teams define their own thresholds for connection drops and request rate anomalies

Verdict

Netdata leads this list because it is the only tool that combines per-second stub_status collection, zero-configuration auto-discovery, dedicated NGINX Plus and VTS collectors, and ML anomaly detection out of the box. Connection spikes and request microbursts that a 15s or 1m poller averages away are visible here at per-second resolution. The honest caveat: the basic integration is intentionally lean, and getting the full Plus API surface means turning on a second collector. For most fleets that trade is minutes of configuration for granularity no polling-based competitor matches by default.

Start free -> See web server monitoring ->

Vendor 02 / 10 · #datadog

Datadog

SaaS observability platform with the deepest NGINX integration among commercial tools: stub_status, full Plus API metrics, logs, and APM tracing.

Best for

Teams already standardized on Datadog for APM, logs, and infrastructure
Organizations running NGINX Plus that want the full Plus API metric set (zones, upstreams, cache, SSL)
Cloud-native shops monitoring NGINX ingress controllers in Kubernetes via Autodiscovery

Pricing

Per-host for infrastructure monitoring; the bill grows with host count
Logs billed per GB ingested and indexed; custom metrics and APM hosts billed separately
Each product (APM, logs, security, RUM) adds its own line item, so cost scales with every capability you turn on

Pros

NGINX integration collects 180+ metrics including server zones, upstream peers, cache hit/miss, SSL handshakes, and limit_req/limit_conn counters
Supports open-source NGINX (stub_status), NGINX Plus (Plus API), and the vhost_traffic_status module
Kubernetes Autodiscovery monitors NGINX ingress controllers via annotations, Helm, or Operator
Correlates NGINX metrics with APM traces to backend services for end-to-end request visibility

Cons

The bill grows with every product added: hosts, log volume, custom metrics, and APM each bill separately, which gets expensive on high-traffic NGINX fleets
The default NGINX log format lacks request response time; you must modify log_format to capture $request_time
15-second default collection interval misses sub-minute spikes, and advanced Plus metrics require configuring status_zone and Plus API endpoints

Verdict

Datadog has the most complete commercial NGINX integration we evaluated, and it earns rank 2 on metric depth alone. If you already pay for Datadog APM, adding NGINX is an obvious move. The trade-offs are granularity and cost shape: the 15s default interval smooths over microbursts, and usage-based billing means a traffic spike hits both your infrastructure and your invoice.

Read full Netdata vs Datadog -> Datadog pricing ->

Vendor 03 / 10 · #prometheus-grafana

Prometheus + Grafana

The de facto open-source stack: Prometheus scrapes NGINX via the official nginx-prometheus-exporter, Grafana visualizes the results.

Best for

Teams with Prometheus expertise who want full control over dashboards and alert rules
Organizations that prefer open-source, vendor-neutral stacks with no per-host licensing
Kubernetes-centric shops already running the Prometheus stack for ingress controller metrics

Pricing

Open source and self-hosted: you run and operate Prometheus, Grafana, Alertmanager, and the exporter yourself
Grafana Cloud is a separate usage-based SaaS option billed per active series and log volume

Pros

Official nginx-prometheus-exporter, maintained by NGINX Inc under Apache 2.0, supports both stub_status and the Plus API
Plus metrics cover server zones, upstream peers, cache, SSL, resolver, rate limiting, and worker stats
PromQL enables custom alert rules and queries across NGINX and infrastructure metrics
Grafana’s ecosystem offers thousands of community dashboards

Cons

Multiple components to assemble and operate: exporter, Prometheus, Alertmanager, Grafana, and storage
No built-in correlation between NGINX logs and metrics; log analysis needs a separate stack like Loki or ELK
15-second default scrape interval misses sub-minute spikes
Scaling to many NGINX instances requires sharding or a layer like Thanos

Verdict

This stack gives you complete NGINX metric coverage through the official exporter and total control over how you use it, at the price of assembling and operating four or five components yourself. It is the right answer for teams with existing Prometheus skills and a Kubernetes-first estate. It is the wrong answer if you want visibility in minutes or per-second resolution, and self-hosted does not mean zero cost - someone owns the uptime of the monitoring stack itself.

Read full Netdata vs Prometheus -> Grafana pricing ->

Vendor 04 / 10 · #newrelic

New Relic

Full-stack observability SaaS with an NGINX integration covering stub_status and Plus API metrics, logs, and inventory data.

Best for

Teams using New Relic for APM that want NGINX metrics alongside application traces
Organizations running both NGINX OSS and NGINX Plus that want one agent for both
Buyers who want pre-built NGINX dashboards without assembling a stack

Pricing

Usage-based: per GB of data ingest, plus per-user pricing by user type
The bill grows with data volume, user count, and retention requirements

Pros

One integration compatible with both NGINX Open Source and NGINX Plus
STATUS_MODULE auto-discovery supports stub_status, ngx_http_status_module, and ngx_http_api_module
Collects metrics, nginx.conf inventory, and logs together, with a pre-built NGINX Monitor quickstart dashboard
Correlates NGINX metrics with APM traces and infrastructure data

Cons

Usage-based pricing makes the bill hard to predict as log and metric volume grows
The example configuration uses a 30-second metric interval, coarser than per-second tools
SaaS-only, with no self-hosted option for data sovereignty requirements

Verdict

New Relic’s NGINX integration is solid and genuinely convenient if you already run its APM: metrics, inventory, and logs from one agent, plus a quickstart dashboard. It sits below Datadog on Plus API depth and below the per-second tools on granularity, and the per-GB model means a busy NGINX fleet is a growing invoice. A reasonable middle choice for existing New Relic shops.

Read full Netdata vs New Relic -> New Relic pricing ->

Vendor 05 / 10 · #dynatrace

Dynatrace

AI-powered observability platform whose OneAgent automatically instruments NGINX and traces processed web requests without manual plugin setup.

Best for

Enterprises wanting automatic discovery of NGINX and surrounding infrastructure from one agent
Teams that value AI-driven root cause analysis across the full stack
Organizations needing on-premises or managed deployment options

Pricing

Per-host pricing with per-GiB data allowances for infrastructure and full-stack monitoring
The bill grows with host count and data volume; additional products bill separately

Pros

OneAgent automatically instruments NGINX code modules and processes web requests without manual configuration
NGINX Plus monitoring plugin queries the Plus status API for upstream and zone metrics
Automatic topology discovery maps NGINX to backend services
Davis AI correlates NGINX anomalies with application and infrastructure causes

Cons

Per-host pricing with data allowances makes large fleets expensive
Independent reviews note missing SSL offloading metrics and unclear documentation for advanced NGINX features
Full value requires adopting the broader Dynatrace platform, not just NGINX monitoring

Verdict

Dynatrace is the strongest option here for automatic, full-stack context: OneAgent finds NGINX, traces requests through it, and lets Davis AI point at the cause. The NGINX-specific metric depth, though, depends on the Plus plugin and configuration, and per-host pricing with data allowances punishes big fleets. Best for enterprises already bought into the platform.

Read full Netdata vs Dynatrace -> Dynatrace pricing ->

Vendor 06 / 10 · #zabbix

Zabbix

Mature open-source monitoring platform with official templates for NGINX stub_status and the full NGINX Plus API.

Best for

Organizations that want a mature, self-hosted platform with no per-host licensing
Teams already running Zabbix for infrastructure that want official NGINX templates
Enterprises needing distributed monitoring with proxies and high availability

Pricing

Open source and self-hosted: you operate the Zabbix server, database, and frontend
Optional commercial support and Zabbix Cloud (SaaS) available

Pros

Official templates: Nginx by Zabbix agent, Nginx by HTTP, and NGINX Plus by HTTP
The Plus template collects zones, upstreams, resolvers, SSL, and rate limiting via the Plus API
Templates ship with pre-built triggers for connection drops, service down, and response time
Supports Zabbix versions 5.0 through 7.4, fully self-hosted with no per-node licensing

Cons

Polling-based architecture with default template intervals around 1 minute misses sub-minute spikes
The Nginx by Zabbix agent template does not support HTTPS or redirects (a web.page.get limitation)
No built-in NGINX access/error log analysis
Requires manual template linking, macro configuration, and ongoing server maintenance

Verdict

Zabbix offers surprisingly complete NGINX coverage for a self-hosted, open-source platform, including full Plus API metrics and pre-built triggers. The 1-minute polling default is its real ceiling for NGINX work: connection spikes that resolve in seconds never appear. A good fit for established Zabbix shops; a heavy lift for anyone starting fresh.

Read full Netdata vs Zabbix -> Zabbix pricing ->

Vendor 07 / 10 · #checkmk

Checkmk

Self-hosted monitoring platform with an nginx_status agent plugin that parses stub_status output for core NGINX performance indicators.

Best for

Teams already using Checkmk for infrastructure monitoring
Organizations preferring agent-based, self-hosted collection
SMBs and enterprises wanting one platform for servers, networks, and applications

Pricing

Raw edition is open source and self-hosted; you operate it yourself
Commercial editions bill per node, so the bill grows with node count

Pros

nginx_status agent plugin auto-detects running NGINX servers and parses stub_status output
Graphs and alerts for active connections, requests, and connection states
Agent-based collection with no external scripts required
Integrates NGINX status with host CPU, memory, and disk monitoring in the same view

Cons

Monitors the global NGINX process only, not per-server or per-zone metrics
No NGINX Plus API support in the standard plugin; Plus metrics need community plugins
No built-in NGINX access/error log analysis
Manual plugin installation into the agent plugins directory

Verdict

Checkmk covers the stub_status basics cleanly and folds them into a capable infrastructure platform, which is all many Checkmk shops need. It is not a deep NGINX tool: no Plus API, no zones, no logs. Treat it as infrastructure monitoring that happens to include NGINX counters, not as NGINX observability.

Read full Netdata vs Checkmk -> Checkmk pricing ->

Vendor 08 / 10 · #site24x7

Site24x7

Cloud-based observability platform with dedicated NGINX and NGINX Plus plugins for connection, request, and SSL metrics.

Best for

SMBs that want a cloud SaaS platform with quick NGINX plugin setup
Teams needing NGINX Plus metrics (handshakes, responses by code, zone data) without building exporters
Organizations wanting website, server, and NGINX monitoring in one product

Pricing

Per-monitor subscription tiers; each monitored NGINX host consumes a monitor license
The bill grows with monitor count and add-on products

Pros

Dedicated NGINX and NGINX Plus plugins covering connections, requests per second, SSL handshakes, and responses by HTTP status code
Plus plugin reads the Plus API for zone-level backend metrics
Automatic AppLogs integration for NGINX access logs
100+ ready-to-use plugins for the rest of the stack, with threshold alerts

Cons

Per-monitor licensing means costs grow as you add NGINX instances
Plugin model requires installing the agent and configuring status URLs manually per host
SaaS-only, with no self-hosted option

Verdict

Site24x7 is a pragmatic SMB choice: real NGINX and Plus plugins, log integration, and alerting without assembling anything. The friction is operational and financial: per-host agent installation and per-monitor licensing both scale linearly with your fleet, and collection granularity stays at typical plugin intervals rather than per-second.

Read full Netdata vs Site24x7 -> Site24x7 pricing ->

Vendor 09 / 10 · #betterstack

Better Stack

Unified observability SaaS that collects NGINX logs via Vector, shows stub_status metrics on a pre-built dashboard, and adds uptime and incident management.

Best for

Teams that want NGINX access/error log analysis as the primary monitoring surface
Organizations wanting uptime monitoring, log management, and incident response in one tool
Buyers who prefer SQL-based log querying over metric-only dashboards

Pricing

Usage-based bundles: per GB of logs, traces, and metrics ingested and retained
Per-monitor for uptime checks and per-responder for incident management
The bill grows with data volume, monitor count, and team size

Pros

One-command Vector setup for NGINX logs with automatic parsing into structured JSON
Pre-built NGINX dashboard with uptime, requests per second, and active connections
SQL-based log querying over ClickHouse for fast troubleshooting
Combines logs, metrics, uptime monitoring, and incident management, with anomaly detection alerts

Cons

Metrics limited to stub_status basics; no Plus API support for zones, upstreams, or cache
Requires running the Vector agent alongside NGINX for log shipping
Usage-based pricing scales with log volume from high-traffic NGINX servers

Verdict

Better Stack is the best tool on this list for NGINX log analysis specifically, and the uptime-plus-incident bundle is genuinely useful for small teams. As an NGINX metrics tool it is shallow: stub_status counters only, no Plus API, no infrastructure context. Pair it with a metrics platform or use it when logs are the question you are actually asking.

Read full Netdata vs Better Stack -> Better Stack pricing ->

Vendor 10 / 10 · #coroot

Coroot

Open-source, eBPF-based observability platform that captures NGINX HTTP traffic, metrics, logs, and traces without code changes.

Best for

Kubernetes-centric teams that want eBPF-based service monitoring without instrumentation
Organizations preferring open-source observability with AI-guided root cause analysis
Teams wanting traces, metrics, logs, and profiles from one agent

Pricing

Open source and self-hosted: you run and operate the Coroot server and node agents
Commercial support and managed options available

Pros

eBPF-based coroot-node-agent captures NGINX HTTP requests without code changes or stub_status configuration
Combines metrics, logs, traces, and continuous profiling in one platform
AI-powered root cause analysis with predefined inspections and SLO-based alerting
Open source with an active project, and event-driven sub-second data capture

Cons

Not NGINX-specific: no stub_status or Plus API collection, so no connection states, cache hit rates, or upstream health counters
Younger project with a smaller ecosystem than established platforms
Requires Linux kernel eBPF support; limited on older kernels or non-Linux hosts
Self-hosted operation means running and maintaining the Coroot stack yourself

Verdict

Coroot takes a genuinely different angle: watch the traffic, not the daemon. For request-level visibility and tracing in Kubernetes that works well and needs zero NGINX configuration. But it cannot tell you about NGINX internals like worker connection states or cache behavior, because it never asks NGINX. Rank it as a complement to NGINX-specific monitoring, not a replacement.

Read full Netdata vs Coroot -> Coroot website ->

Frequently Asked Questions

What NGINX metrics should I monitor?

Start with active connections and connection states (reading, writing, waiting), plus requests per second and accepted/handled connection rates. Track HTTP 4xx and 5xx error rates, upstream server health with response times and failed requests, and SSL handshake failures and session reuses. If you use NGINX as a cache, watch hit/miss ratios. At the OS level, file descriptor usage and connection backlog saturation are the two silent killers: they cap concurrency long before CPU or memory do.

How do I enable NGINX stub_status for monitoring?

Add a location block to nginx.conf: location /stub_status { stub_status; }, restricted with allow/deny directives so only localhost or your monitoring agent’s IP can reach it. Reload with nginx -s reload or systemctl reload nginx, then verify with curl http://localhost/stub_status. Netdata auto-detects the endpoint with zero configuration; other tools require you to point them at the status URL explicitly.

What is the difference between monitoring NGINX open source and NGINX Plus?

NGINX OSS exposes only the stub_status counters: active connections, accepted/handled, requests, and reading/writing/waiting. NGINX Plus exposes a much richer API with server zones, upstream peers, cache, SSL, resolver, and rate limiting metrics. Tool support differs: Datadog, New Relic, Zabbix, and Site24x7 have dedicated Plus integrations, Netdata ships a separate NGINX Plus collector, and the official nginx-prometheus-exporter supports both. Better Stack and Checkmk do not collect Plus API metrics.

Can I monitor NGINX logs and metrics together?

Yes, and you should. Access logs show request latency, status codes, and client behavior; error logs show upstream failures and configuration issues. Netdata correlates systemd-journal logs with metrics, while Datadog and Better Stack parse NGINX access and error logs into structured data. Datadog recommends adding $request_time to the log_format to capture response times. Prometheus + Grafana has no native answer here: you need a separate log stack such as Loki or ELK.

What is the best free or open-source NGINX monitoring tool?

The Netdata agent is open source (AGPL) with per-second collection, auto-discovery, and a free Cloud tier for small fleets. Prometheus + Grafana with the official nginx-prometheus-exporter is fully open source but requires assembling multiple components. Zabbix ships official NGINX templates for OSS and Plus, and Checkmk Raw edition includes the nginx_status plugin. Remember that open source means you run and operate it: the software carries no license fee, but the operating cost is real.

How often should I collect NGINX metrics?

Per-second collection catches connection spikes and request microbursts that minute-averaged monitoring misses entirely. Prometheus defaults to 15-second scrapes and Zabbix templates to 1-minute polling; Netdata collects per second by default. Short-lived issues like connection backlog saturation or TLS CPU saturation can appear and resolve between 15s or 1m polls, leaving you with a clean dashboard and a real outage.

How do I monitor NGINX in Kubernetes?

The NGINX ingress controller exposes Prometheus metrics on port 10254 by default, and the nginx-prometheus-exporter has specific ingress controller documentation. Datadog Autodiscovery monitors the ingress via annotations, Helm, or Operator. Netdata provides a Helm chart with auto-discovery of pods, containers, and ingress controllers. Whichever tool you pick, monitor both the ingress controller pods and the upstream services behind NGINX, because most ‘NGINX is slow’ reports are actually slow backends.

What causes high NGINX response times?

Slow upstream application backends are the most common cause by far. After that: connection backlog saturation when active connections exceed worker limits, file descriptor exhaustion (ulimit) capping simultaneous connections, TLS CPU saturation from expensive handshakes without session resumption, and buffer spills to disk when proxy_buffers are too small. Tools that track upstream response times, such as Datadog, the Netdata NGINX Plus collector, or the nginx-prometheus-exporter, help isolate whether NGINX or the backend is the slow part.

How much does NGINX monitoring cost?

NGINX monitoring ranges from free open-source tools to commercial platforms starting at $4.50/node/month. Per-node models like Netdata’s start at $4.50/node/month on annual plans with unlimited metrics and logs. Per-host models (Datadog, Dynatrace) bill per host with each additional product priced separately. Usage-based models (New Relic, Better Stack) bill per GB ingested, so the bill tracks your traffic. Open-source options (Prometheus + Grafana, Zabbix, Checkmk Raw) have no license cost but you operate the infrastructure. The practical question is which growth curve—host count, data volume, monitor count, or node count—hurts least as your fleet scales.

The best NGINX monitoring tools, ranked

How we evaluated NGINX monitoring tools

Netdata

Datadog

Prometheus + Grafana

New Relic

Dynatrace

Zabbix

Checkmk

Site24x7

Better Stack

Coroot

Frequently Asked Questions

Book Your Free Demo