Buyer’s Guide - August 2026

The 9 best NATS monitoring tools, ranked

Q: How do I monitor NATS JetStream streams and consumers?

You monitor NATS JetStream streams and consumers by scraping the /jsz HTTP monitoring endpoint and consuming system-account advisories. Tools like Netdata’s NATS collector, prometheus-nats-exporter, NATS Surveyor, and Elastic Metricbeat can collect these metrics, though some require explicit configuration for per-account, per-stream, and per-consumer details.

Q: What metrics should I track for NATS?

You should track NATS metrics across server, JetStream, connection, and cluster levels, including in/out bytes per second, active connections, stream storage used, pending bytes as a slow-consumer indicator, and route traffic. Per-account metrics are also essential if you use NATS accounts for tenancy.

Q: Prometheus exporter vs NATS Surveyor: which should I use?

Choose prometheus-nats-exporter if you prefer a per-server HTTP scraper, or choose NATS Surveyor if you want a single system-account connection to observe the whole deployment. Both require Prometheus plus Grafana for storage and dashboards. Surveyor requires system accounts enabled and credentials, while the exporter only needs the HTTP monitoring port.

Q: How do I monitor NATS in Kubernetes?

To monitor NATS in Kubernetes, deploy prometheus-nats-exporter as a sidecar or separate pod and scrape it with the Prometheus Operator. Alternatively, NATS Surveyor can run as a single deployment using system account credentials, or Netdata can auto-discover NATS containers to collect metrics alongside pod and node data.

Q: What is the NATS monitoring port?

The default NATS HTTP monitoring port is 8222, configured via the http_port or https_port server option. The monitoring server is separate from the client port (default 4222) and should not be exposed publicly. Collectors like Netdata auto-detect NATS on localhost:8222. The /healthz endpoint is used for health checks and can verify that JetStream is enabled.

Q: How do I alert on NATS slow consumers?

To alert on NATS slow consumers, create an alert based on the slow_consumers counter from the varz endpoint or accumulated pending bytes on a connection. You can set this up using tools like Netdata, Prometheus with PromQL, or Synadia Insights, which includes slow-consumer detection among its 100+ built-in checks.

Q: What is the difference between NATS monitoring and JetStream monitoring?

Core NATS monitoring covers server traffic, messages, connections, routes, and subscriptions, while JetStream monitoring specifically tracks stream and consumer state like storage used, messages stored, consumer lag, and API request rates and errors. Modern deployments need both, as tools that only cover core NATS will miss critical JetStream health data.

NATS is not a generic service you can point any APM agent at. Server stats, JetStream streams and consumers, routes, gateways, and leaf nodes all live behind NATS’s own HTTP monitoring API and system-account event stream, and only NATS-aware collectors exploit them. We ranked nine tools on metric depth, collection resolution, alerting, and total cost for production NATS fleets.

Start with Netdata (free) Jump to rankings

Why this list exists

NATS monitoring is a distinct problem from general infrastructure monitoring. nats-server exposes a rich HTTP monitoring API on port 8222 (varz, connz, routez, subsz, gatewayz, leafz, healthz, jsz, accountz) plus a system-account event stream. A collector that does not understand those endpoints gives you a fraction of the picture. The most common buying mistake is assuming a general-purpose APM or a default Prometheus scrape delivers NATS depth. In practice, most general tools stop at varz-level counters, while JetStream stream and consumer health, per-account usage, and gateway and leaf-node topology need a collector built for NATS semantics.

Three dimensions decide most outcomes:

Metric depth. Does the tool cover more than basic varz counters? JetStream streams and consumers, per-account metrics, routes, gateways, leaf nodes, and connection-level pending bytes are what separate a real NATS monitor from a checkbox integration.
Collection resolution. NATS is a high-throughput messaging system. Message-rate spikes and slow-consumer buildup happen in seconds, so a 10 to 15 second polling interval routinely misses the incident you are trying to explain.
Time to value and cost shape. Auto-detection and prebuilt content beat a stack you wire together yourself, and per-host pricing punishes the many-small-nodes deployment pattern NATS fleets tend to follow.

A note on pricing: we do not quote competitor list prices in this guide. List prices for usage-based platforms are almost meaningless without knowing your ingestion volume, retention, and seat count, and quoting them out of context misleads more than it informs. Instead, each card describes the pricing shape and what makes the bill grow, with a link to the vendor’s own pricing page.

If you are still building your runbooks, our NATS monitoring guides walk through the operational side: which endpoints matter, how to read slow-consumer signals, and how to structure alerts for JetStream.

We assembled the shortlist from tools with verified, documented NATS support: official NATS project utilities, dedicated collectors, and commercial platforms with a named NATS integration. Vendors with only a passing mention of NATS in a generic integrations catalog were excluded.

Weighting reflects what actually decides a NATS monitoring purchase. Metric depth carries the most weight because shallow coverage is the failure mode buyers hit first. Collection resolution is second because NATS incidents are fast. Alerting, time to value, and cost efficiency share the remainder, with ecosystem integration as a tiebreaker.

Tester credit

Compiled by the Netdata team - Updated August 12, 2026

Sources verified

Scoring criteria

NATS metric depth 25%
JetStream, accounts, routes, gateways, leaf nodes, connection detail
Collection resolution 20%
Per-second vs 10-15s polling for fast-moving message rates
Time to value 15%
Auto-detection and prebuilt content vs assembled stacks
Alerting and anomaly detection 15%
NATS-specific conditions, slow consumers, JetStream API errors
Cost efficiency 15%
Pricing shape vs fleet size; operational cost of self-hosting
Ecosystem integration 10%
Prometheus compatibility, Grafana, Kubernetes, alert routing

Vendor 01 / 09 · #netdata

Netdata

An open-source, per-second monitoring platform with a dedicated NATS collector covering server, JetStream, accounts, routes, gateways, and leaf nodes.

Best for

Teams running NATS across many servers or edge nodes who want per-second visibility without per-GB ingestion costs
Operators who want zero-config auto-detection of NATS instances on localhost or in Docker
Platform teams that want NATS metrics alongside host-level CPU, memory, and network context in one dashboard

Pricing

Per-node pricing: Netdata Cloud Business starts at $4.5/node/month on annual plans, with the per-node price decreasing as node count grows
Agents are open source (AGPL) and free to run
Free Cloud tier exists for small fleets
Bill grows with monitored node count, not with metrics volume or retention

Pros

Dedicated NATS collector (go.d.plugin nats module) with a 1-second default collection interval
Covers server traffic, messages, connections, CPU/memory/uptime, health probe status, JetStream streams, consumers, storage, API requests and errors, per-account metrics, routes, gateways, and leaf-node connections with RTT
Auto-detects NATS on localhost port 8222 and inside Docker containers with zero configuration
Supports multiple instances including remote servers, HTTP auth, TLS, and proxies
Machine-learning anomaly detection on collected metrics with no manual threshold tuning
800+ integrations total, so NATS charts sit next to host, container, and application data

Where teams pair it

No alert rules ship preconfigured for the NATS collector; you create alerts for slow consumers and JetStream API errors yourself
No dedicated prebuilt NATS dashboard; you compose dashboards from the collected NATS charts
No NATS client application tracing; teams needing end-to-end request tracing pair it with an APM tool

Verdict

Netdata leads this list because it is the only option that combines the deepest open-source NATS metric coverage with a 1-second collection interval in a single package. Server, JetStream, account, route, gateway, and leaf-node scopes are all collected, and auto-detection on port 8222 means the agent finds NATS without configuration. The per-node pricing model keeps cost predictable as fleets grow across many small nodes, where per-host and per-GB models punish you. The honest gaps: no prebuilt NATS alert rules, no dedicated NATS dashboard out of the box, and the NATS HTTP monitoring port must be enabled, a prerequisite shared by every HTTP-based collector on this list.

Start free -> See the NATS integration ->

Vendor 02 / 09 · #prometheus-grafana

Prometheus + Grafana

The de-facto open-source stack for NATS, pairing the official prometheus-nats-exporter with Prometheus storage and Grafana dashboards.

Best for

Teams already standardized on Prometheus and Grafana who want NATS in existing dashboards
Organizations that prefer assembling best-of-breed open-source components
Kubernetes-centric deployments where Prometheus is already the cluster standard

Pricing

Open source and self-hosted: you run and operate Prometheus, Grafana, and the exporter yourself
Grafana Cloud offers usage-based SaaS tiers with free and paid plans
Cost grows with metric cardinality, retention, and the operational effort of running the stack

Pros

Official prometheus-nats-exporter maintained by nats-io (Apache-2.0, active as of July 2026) aggregates varz, connz, subz, routez, healthz, and jsz endpoints
Exporter supports detailed connection metrics, account stats, gateway, leaf, and JetStream stream/consumer metrics via flags
Community Grafana dashboards exist for NATS servers and JetStream (dashboard IDs 2279 and 14725)
PromQL gives flexible alerting on any NATS metric
Huge ecosystem: Alertmanager, Thanos, Grafana Cloud, and hundreds of compatible tools

Cons

Requires assembling and operating exporter, Prometheus, and Grafana as separate components
Default scrape intervals of 10-15s miss sub-second NATS traffic spikes unless tuned
JetStream metrics create high cardinality in Prometheus; operators must filter jsz metrics carefully
No NATS-specific anomaly detection; alerting means writing manual PromQL thresholds

Verdict

This is the standard community answer for NATS monitoring, and for good reason: the official exporter covers nearly every NATS monitoring endpoint, and the ecosystem around it is unmatched. It ranks below Netdata because you are assembling three components, and default scrape intervals are too coarse for NATS traffic patterns. If your organization already runs Prometheus well, this is the path of least resistance. If you do not, budget real operational time for it. Open source here means self-hosted: the license is free, the operations are not.

Read full Netdata vs Prometheus -> Grafana pricing ->

Vendor 03 / 09 · #nats-surveyor

NATS Surveyor

A NATS-native Prometheus exporter from the NATS team that observes an entire deployment from a single system-account connection.

Best for

NATS operators who want one exporter to see the whole cluster without per-server sidecars
Teams on NATS 2.0+ system accounts who want per-account, per-route, and JetStream visibility
Synadia platform users who want a lighter open-source alternative to Synadia Insights

Pricing

Open source (Apache-2.0) and self-hosted: you operate Surveyor plus Prometheus and Grafana
Docker Compose stack included for quick start
Cost is operational: running the exporter, Prometheus storage, and Grafana

Pros

Maintained by nats-io (Apache-2.0, active as of August 2026); used extensively by Synadia
Single exporter connects to one NATS server and discovers the entire deployment via system-account Statz messages
Per-account metrics, gateway metrics, Raft group metrics, and JetStream stream/consumer metrics with leader-only and filter options to control cardinality
Ships with a docker-compose stack (Surveyor + Prometheus + Grafana) and a prebuilt dashboard
Provides a nats_up metric to distinguish exporter connectivity problems from NATS problems

Cons

Requires NATS system account credentials and system accounts enabled on the server
Requires Prometheus and Grafana (or another Prometheus-compatible stack) for storage and visualization
JetStream metrics can cause high cardinality in Prometheus unless jsz filters and leader-only mode are used
No built-in alerting or anomaly detection; relies on Alertmanager rules

Verdict

Surveyor is the most NATS-native open-source exporter, designed specifically for whole-deployment visibility through the system account. One connection gives you per-account, gateway, Raft, and JetStream data without an exporter per server, which scales better than prometheus-nats-exporter on large clusters. It ranks third because it is still only the collection layer: without Prometheus and Grafana behind it, it shows you nothing. For NATS 2.0+ deployments with system accounts already in place, it is the strongest exporter choice.

NATS Surveyor on GitHub -> NATS monitoring docs ->

Vendor 04 / 09 · #synadia-insights

Synadia Insights

A commercial, NATS-native observability product from the company behind NATS that indexes everything a system account can observe into a queryable time-series graph.

Best for

Production deployments where NATS is critical infrastructure and teams need entity-level diagnosis (which connection, which account, which stream)
Teams that want 100+ built-in NATS checks with severity and remediation guidance without writing their own rules
Organizations that want AI-agent queryable NATS state without granting agents direct system access

Pricing

Annual contract priced per number of NATS servers monitored; a significant commitment for smaller teams
30-day free trial with a built-in simulator for evaluation
Runs as a single read-only binary on a machine of your choosing; nothing installed on NATS nodes

Pros

Purpose-built for NATS by Synadia, the company behind the NATS project
Indexes server state, cluster topology, route health, connection metadata, subject-level message counts, JetStream asset state, account usage, and entity relationships
100+ built-in automated checks with severity, plain-language description, and remediation guidance
Read-only design: cannot edit configurations, cannot access message payloads or stream data
Ships with an AI agent skill so agents can query NATS state in plain English
Complements existing Grafana, Datadog, and OpenTelemetry stacks rather than replacing them

Cons

Requires NATS Server v2.10+ and system account credentials
Annual contract pricing is a significant commitment for smaller teams
Does not replace a general monitoring stack; positioned as a NATS-specific complement
Does not work with Synadia Cloud, since it requires direct system account access

Verdict

This is the deepest NATS-specific diagnostics available commercially. It goes beyond metrics into entity relationships and configuration-change context, and the 100+ built-in checks are the opposite of Netdata’s and Prometheus’s write-your-own-alerts model. It ranks fourth because the annual contract and NATS v2.10+ requirement narrow the audience: for teams where NATS is critical infrastructure, it is compelling; for everyone else, it is more than they need. Note that it complements, rather than replaces, a general metrics platform.

Synadia Insights product page -> Synadia Insights docs ->

Vendor 05 / 09 · #new-relic

New Relic

A commercial observability platform with a NATS quickstart that instruments NATS applications via the Go agent, with a prebuilt dashboard and starter alerts.

Best for

Teams already on New Relic who want NATS visibility alongside APM, logs, and traces in one platform
Organizations that want a prebuilt NATS dashboard and four starter alerts without building from scratch
Application teams that want client-side NATS visibility rather than server-side endpoint scraping

Pricing

Usage-based: per-user seats plus data ingestion volume
A free tier is available with ingestion limits for small-scale evaluation
Bill grows with ingestion volume, retention, and full-platform user count

Pros

Official NATS quickstart built by New Relic with one dashboard, four alerts, and documentation
Starter alerts cover memory above 90%, transaction errors above 10%, CPU above 90%, and Apdex below 0.5
Go agent instrumentation gives transaction tracing, error inbox, and service maps for NATS-based applications
Deep platform integration with logs, traces, and infrastructure monitoring

Cons

Quickstart is application-centric (Go agent) and does not scrape NATS HTTP monitoring endpoints for server or JetStream metrics
No JetStream stream or consumer metrics in the quickstart
Usage-based pricing can grow quickly with ingestion volume across a large NATS fleet
Requires embedding the New Relic Go agent in each NATS client application

Verdict

New Relic’s NATS support is a different thing from what most of this list offers: it watches your NATS client applications, not the NATS servers. That is genuinely useful for application teams who want tracing and service maps, and the quickstart is the fastest prebuilt-alert experience here. But for the operators reading this guide, the absence of server-side varz scraping and any JetStream coverage is disqualifying as a primary NATS monitor. Treat it as an APM complement, not a NATS monitoring solution.

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

Vendor 06 / 09 · #elastic

Elastic Observability

Elastic’s Metricbeat NATS module scrapes stats, connections, routes, subscriptions, and optional JetStream metrics into Elasticsearch and Kibana.

Best for

Teams already running Elasticsearch and Kibana who want NATS metrics alongside logs and APM
Organizations that want a single platform for NATS metrics, NATS logs, and application traces
Teams that prefer the Elastic Agent and Metricbeat unified collection model

Pricing

Usage-based SaaS (Elastic Cloud) priced on ingestion volume and retention
Open source self-managed option: you run and operate Elasticsearch, Kibana, and Metricbeat
Bill grows with data volume, not node count

Pros

Official Metricbeat NATS module with default metricsets: stats, connections, routes, subscriptions
Optional connection, route, and jetstream metricsets for per-stream and per-consumer detail
Tested with NATS 2.2.6 and 2.11.x
Ships with a predefined Kibana dashboard for NATS
Supports TLS, HTTP auth, and standard Metricbeat configuration options

Cons

Documented example uses a 10s collection period, coarser than per-second monitoring
JetStream metricset requires explicit per-account, per-stream, and per-consumer name configuration
Running Elasticsearch and Kibana is operationally heavy for NATS monitoring alone
No NATS-specific anomaly detection; relies on Elastic ML jobs or manual Kibana alerts

Verdict

The Metricbeat module is solid server-side coverage, and it is one of the few commercial integrations with a real JetStream metricset. Two things hold it back for NATS-first buyers: the 10s example period is too coarse for messaging traffic, and the JetStream metricset needs you to enumerate accounts, streams, and consumers by name, which does not scale with dynamic workloads. If your organization already runs Elastic, enabling the module is an easy win. Standing up the Elastic stack just for NATS is not.

Read full Netdata vs Elastic -> Elastic pricing ->

Vendor 07 / 09 · #datadog

Datadog

Datadog’s NATS support is a community-maintained gnatsd check in integrations-extras collecting 25 server, connection, and route metrics.

Best for

Teams already standardized on Datadog who want basic NATS server visibility in existing dashboards
Organizations that accept community-tier integration support for NATS
Teams monitoring NATS alongside a large commercial SaaS infrastructure stack

Pricing

Per-host pricing plus per-GB log and metric ingestion
Community integration is free to install but requires a paid Datadog subscription
Bill grows with host count and ingestion volume

Pros

Collects 25 metrics across varz, connz, and routez: traffic, messages, connections, subscriptions, slow consumers, memory, routes, remotes
Service check gnatsd.can_connect reports CRITICAL when the Agent cannot reach the endpoint
Metrics tagged with cluster names for multi-cluster environments
Compatible with all major platforms

Cons

Community integration: not included in the Agent package, installed separately via datadog-agent integration install
No JetStream metrics at all; no account, gateway, or leaf-node coverage
No NATS-specific events or prebuilt dashboards
Community integrations get best-effort support rather than Datadog’s standard SLA

Verdict

Datadog’s NATS check is the weakest coverage-per-dollar on this list. Twenty-five varz, connz, and routez metrics is basic server health, and the complete absence of JetStream support is a serious gap for any modern NATS deployment where streams and consumers are the product. The community-tier status means best-effort support and a manual install step. If you are already deep in Datadog and only need to know the server is alive and how busy it is, it works. If JetStream matters to you, look elsewhere or pair it with a NATS-aware collector.

Read full Netdata vs Datadog -> Datadog pricing ->

Vendor 08 / 09 · #instana

IBM Instana

IBM Instana provides automatic NATS monitoring with application discovery, service mapping, tracing, and automated troubleshooting inside its APM platform.

Best for

Enterprises already using Instana for APM who want NATS included in automatic discovery
Teams that value automatic instrumentation over manual exporter configuration
Organizations running NATS inside a larger microservices estate with tracing needs

Pricing

Per-host per-month pricing, billed annually across Essentials and Standard tiers
Self-hosted option priced per managed virtual server
Bill grows with the number of monitored hosts

Pros

NATS is a listed supported technology with automatic application discovery and service mapping
Provides tracing and automated troubleshooting for NATS-based applications
Agent-based automatic instrumentation requires no manual exporter setup
Part of a broad APM platform covering hundreds of technologies

Cons

NATS monitoring is application-centric (tracing and service maps) rather than server-infrastructure-centric
No documented JetStream stream or consumer metrics collection
Per-host pricing is expensive for large NATS fleets spread across many small nodes
NATS is one of hundreds of supported technologies; NATS-specific depth is limited compared to NATS-native tools

Verdict

Instana’s NATS story mirrors New Relic’s: it is about the applications talking to NATS, not the NATS servers themselves. Automatic discovery and service mapping are real conveniences, and if Instana is already your APM, NATS comes along for free in operational terms. But there is no documented JetStream metric collection, and per-host pricing stacks up painfully on the many-small-nodes topology NATS fleets favor. It ranks eighth because it answers a different question than the one this guide is about.

Read full Netdata vs Instana -> IBM Instana pricing ->

Vendor 09 / 09 · #nats-top

nats-top

A top-like terminal tool from the NATS team for live inspection of server stats and per-connection activity.

Best for

Operators who want an instant terminal view of a NATS server without installing a platform
Quick debugging: connection counts, message rates, slow consumers, per-connection stats
Scripted one-shot snapshots via the -o flag

Pricing

Open source (MIT) and self-hosted: build or download the binary and run it
No ongoing cost; operational footprint is a single CLI tool

Pros

Maintained by nats-io (MIT license, active as of June 2026)
Shows server load (CPU, memory, slow consumers), in/out message and byte rates, and per-connection detail including subscriptions, pending bytes, and uptime
Interactive sorting by connection ID, subscriptions, messages, bytes, or idle time
Supports TLS, basic auth, and HTTPS monitoring endpoints
Installable via curl script, go install, or RPM/Deb packages

Cons

Terminal-only: no dashboards, no history, no alerting
No JetStream, account, gateway, or leaf-node metrics
Single-server view; does not aggregate a cluster
A debugging utility, not a monitoring platform

Verdict

nats-top earns its place as the tool you reach for in the first sixty seconds of an incident: is the server loaded, which connection is backing up, are slow consumers accumulating. The 1-second refresh and per-connection pending bytes make it genuinely good at that job. But it keeps no history, fires no alerts, and sees one server at a time, so it cannot be your monitoring answer. Install it everywhere anyway; it pairs well with every platform on this list.

nats-top on GitHub -> NATS monitoring docs ->

Frequently asked questions

How do I monitor NATS JetStream streams and consumers?

What metrics should I track for NATS?

Does NATS have built-in monitoring?

Yes. nats-server provides a lightweight HTTP monitoring server on a dedicated port (default 8222) with endpoints including /varz, /connz, /routez, /subsz, /gatewayz, /leafz, /healthz, /jsz, and /accountz. These endpoints are what most NATS collectors scrape, including Netdata, prometheus-nats-exporter, Metricbeat, and Datadog’s gnatsd check. NATS also emits system-account events (Statz messages and advisories), which NATS Surveyor and Synadia Insights consume instead of scraping per-server endpoints.

Prometheus exporter vs NATS Surveyor: which should I use?

How do I monitor NATS in Kubernetes?

What is the NATS monitoring port?

How do I alert on NATS slow consumers?

Is there a free NATS monitoring tool?

Yes. Netdata’s agent is open source (AGPL) with a free Cloud tier for small fleets, and the NATS collector is included in all tiers. prometheus-nats-exporter, NATS Surveyor, and nats-top are open source (Apache-2.0 and MIT) but you operate Prometheus and Grafana yourself. Elastic Metricbeat’s NATS module is open source but requires running Elasticsearch and Kibana. Open source means self-hosted: the license is free, but the operational cost of running the stack is real.

What is the difference between NATS monitoring and JetStream monitoring?

The 9 best NATS monitoring tools, ranked

How we evaluated NATS monitoring tools

Netdata

Prometheus + Grafana

NATS Surveyor

Synadia Insights

New Relic

Elastic Observability

Datadog

IBM Instana

nats-top

Frequently asked questions

Book Your Free Demo