Buyer’s Guide - August 2026

The 10 best CockroachDB monitoring tools, ranked by an operator

Q: What are the most important CockroachDB metrics to watch?

Start with SQL service latency (sql.service.latency) and connection counts, node liveness and heartbeats, unavailable and under-replicated ranges, storage capacity and disk usage, and Raft leader and leaseholder counts. Watch clock offset closely: CockroachDB kills nodes that drift past the 400ms limit, so offset alerts are early warnings of node loss, not noise.

Q: Does CockroachDB have built-in monitoring?

Yes. The DB Console provides time-series charts for SQL queries, jobs, and node health, and CockroachDB ships built-in safeguards and alerting rules. But those rules need external delivery, and CockroachDB recommends exporting metrics to an external tool for production alerting, retention, and cross-stack correlation.

Q: Can I monitor CockroachDB with Prometheus and Grafana?

Yes, and it is the best-documented path. CockroachDB publishes an official Prometheus scraping guide, Cockroach Labs publishes starter Grafana dashboards, and Grafana Cloud has a native integration with 4 prebuilt dashboards and 8 alerts. Alertmanager handles notification routing. The trade is operational: you run and scale the stack yourself.

Q: What is the best way to alert on CockroachDB issues?

Use a tool with a prebuilt CockroachDB alert pack rather than writing thresholds from scratch: Netdata ships prebuilt CRDB alerts, Grafana Cloud ships 8, and New Relic ships 13. The conditions that matter most are node down, unavailable ranges, low disk, clock offset, and certificate expiry. For Prometheus-based setups, Alertmanager is the standard routing layer.

Q: Can I monitor CockroachDB Cloud with third-party tools?

Yes. CockroachDB Cloud Standard and Advanced tiers expose a Prometheus metrics endpoint that any compatible tool can scrape. Datadog has an official CockroachDB Cloud integration, and Amazon CloudWatch and GCP Logging are supported for metrics and logs export. One exception: DBmarlin supports self-hosted CockroachDB only, per CockroachDB’s documentation.

Every CockroachDB node already speaks Prometheus on /_status/vars, so the real buying question is which tool turns that endpoint into alerts, SQL-level answers, and a bill you can predict. We ranked 10 tools on CockroachDB integration depth, metric resolution, alerting, deployment flexibility, pricing predictability, and SQL visibility. The DB Console alone does not make the cut for production alerting, and neither does a bare scrape config.

Start with Netdata (free) Jump to rankings

The 10 best CockroachDB monitoring tools, ranked by an operator product interface

Why this list exists

CockroachDB monitoring is a different problem from generic database monitoring. Every node exposes a Prometheus-format metrics endpoint at /_status/vars, refreshed roughly every 10 seconds, and the built-in DB Console only shows cluster-internal state. CockroachDB’s own documentation tells you to export metrics to an external system for production alerting and lists Datadog, Prometheus, DBmarlin, CloudWatch, and Kibana as official third-party integrations. Any Prometheus-compatible tool can consume the metrics, but that is where the similarity ends.

The mistake buyers make is one of two assumptions. First, that the DB Console is enough for production. It is not: it has no external alert delivery, limited retention, and no way to correlate a spike in SQL latency with the disk or network problem on the host underneath it. Second, that any Prometheus scraper gives the same depth as a purpose-built CockroachDB integration. It does not. The difference shows up in prebuilt dashboards, in whether the vendor ships CockroachDB-specific alert conditions like unavailable ranges and clock offset, and in whether you can see SQL statement statistics without opening a SQL shell.

Three dimensions decide the outcome of this purchase:

Integration depth and resolution. A dedicated CockroachDB collector or template with prebuilt charts and alerts beats a generic scrape job. And collection interval matters: per-second or 10-second collection catches spikes that a 15 to 60 second Prometheus interval averages away.
Alerting and anomaly detection. Static thresholds on the wrong metric are how you get paged for nothing and miss the under-replicated range that matters. Look for prebuilt CRDB alert packs and, ideally, machine learning on the metrics.
Pricing shape. CockroachDB metrics are high-cardinality. A per-GB or per-active-series bill grows in ways a per-node bill does not.

One note on prices: we do not quote competitor list prices on this page. They change, negotiated discounts vary widely, and a stale number is worse than no number. Instead we describe each vendor’s pricing shape and link the official pricing page from every card. For deeper operational guidance, including the metrics that actually predict trouble, see our CockroachDB monitoring guides.

We built the shortlist from CockroachDB’s official third-party integrations documentation, the monitoring tools CockroachDB users actually deploy, and the platforms most often shortlisted by SRE and platform teams running distributed SQL in production. Every tool here can ingest CockroachDB metrics today; none is included on a roadmap promise.

The criteria are weighted toward what changes your day-to-day: how deep the CockroachDB integration goes and how fast it sees change together carry 45% of the score. Pricing predictability carries 15% because high-cardinality CockroachDB telemetry is exactly the workload that turns a per-series or per-GB bill into a surprise. SQL-level visibility carries 10% because it is the difference between knowing the cluster is slow and knowing which statement made it slow.

Tester credit

Compiled by the Netdata team - Updated August 12, 2026

Sources verified

Scoring criteria

CockroachDB integration depth 25%
Dedicated collector, template, or quickstart vs generic Prometheus scraping
Metric resolution and real-time visibility 20%
Per-second or 10-second collection vs 15-60 second scrape intervals
Alerting and anomaly detection 15%
Prebuilt CRDB alert packs, ML, and AI-assisted root cause
Deployment flexibility and data sovereignty 15%
Self-hosted options and whether telemetry leaves your infrastructure
Pricing predictability 15%
Per-node vs per-GB vs per-active-series on high-cardinality metrics
SQL-level visibility 10%
Statement statistics, query plans, and transaction contention

Vendor 01 / 10 · #netdata

Netdata

Real-time infrastructure and database monitoring with an open-source agent, per-second metrics, ML on every metric, and a dedicated CockroachDB collector.

Best for

Teams that want per-second CockroachDB host context and CRDB metrics without assembling a Prometheus plus Grafana stack
SREs who need ML anomaly detection and AI-assisted root-cause analysis on database metrics
Fleets of any size that want a predictable per-node bill instead of per-GB or per-series metering

Pricing

Per-node pricing: Netdata Cloud Business from $4.50/node/month on annual plans, decreasing as node count grows
Agents are open source (AGPL) and free; a free Cloud tier covers small fleets
Unlimited metrics, logs, users, and retention included; no per-GB or per-series charges
High-cardinality CockroachDB clusters do not change the bill

Pros

Dedicated CockroachDB collector (go.d/cockroachdb) scrapes /_status/vars at a 10-second default, tunable to 1 second
60+ CockroachDB charts covering SQL statements, KV transactions, ranges, replicas, Raft leaders, RocksDB cache, storage capacity, and node liveness
5 built-in CockroachDB alerts: storage capacity, unavailable ranges, under-replicated ranges, and file descriptor limits
top-queries and running-queries functions expose statement statistics from crdb_internal.cluster_statement_statistics and SHOW CLUSTER STATEMENTS
ML anomaly detection on every metric and AI-assisted troubleshooting, with a verified 80% MTTR reduction
800+ integrations, open-source agent, SOC 2 Type 2 certified

Where teams pair it

CRDB metrics land at a 10-second default because CockroachDB only refreshes /_status/vars every 10 seconds; per-second collection applies to host metrics, not CRDB’s own counters
No explain-plan or wait-state analysis at DBmarlin’s depth; teams doing heavy query tuning pair it with a database-specialist tool
No auto-detection for the CockroachDB integration: you configure the endpoint manually, and collector UI configuration requires a paid Cloud plan

Verdict

Netdata is the only tool in this list that pairs a purpose-built CockroachDB collector with per-second host telemetry and machine learning out of the box. You get 60+ CRDB charts, five CRDB-specific alerts, and SQL statement visibility without writing a scrape config or a dashboard. The per-node price stays flat no matter how many series your cluster emits, which matters when CockroachDB’s cardinality meets a per-series competitor bill. The honest caveat: CockroachDB’s endpoint only refreshes every 10 seconds, so CRDB counters are 10-second data even though host metrics are per-second, and if you need plan-level query tuning you will still want DBmarlin alongside it.

Start free -> See database monitoring ->

Vendor 02 / 10 · #datadog

Datadog

Commercial SaaS observability platform with a first-party CockroachDB integration collecting hundreds of cockroachdb.* metrics via the Datadog Agent.

Best for

Teams already standardized on Datadog for APM, infrastructure, and logs
Organizations that want out-of-the-box CockroachDB dashboards and alerting with minimal setup
Enterprises correlating CockroachDB metrics with application traces and logs

Pricing

Per-host pricing for infrastructure, with separate per-host charges for APM and Database Monitoring
Per-GB pricing for log ingestion and per-event pricing for other data types
The bill grows with host count, APM hosts, and log volume; high-cardinality clusters make totals hard to forecast

Pros

First-party OpenMetrics-based integration collecting hundreds of cockroachdb.* metrics across admission control, changefeeds, SQL, storage, Raft, and transactions
Official CockroachDB Cloud integration for Standard and Advanced tiers, plus a self-hosted Agent check
Out-of-the-box dashboards with CRDB metrics available for alerting and correlation
1,000+ integrations and strong APM, infrastructure, and logs correlation

Cons

SaaS-only: your metrics leave your infrastructure
Cost compounds across hosts, APM, and log volume; large high-cardinality fleets see unpredictable bills
No dedicated SQL-level query analysis for CockroachDB beyond the Prometheus-exported metrics

Verdict

Datadog has the deepest first-party CockroachDB integration of any SaaS platform: hundreds of metrics, official CockroachDB Cloud support, and mature alerting on day one. If your organization already pays for Datadog APM and infrastructure, adding CockroachDB is the path of least resistance and the correlation across traces and logs is genuinely good. The trade-off is the bill. Per-host plus per-GB pricing on a high-cardinality database is how teams end up with an observability invoice that rivals the database itself. We do not quote list prices here; check the shape against your fleet size on their pricing page.

Read full Netdata vs Datadog -> Datadog pricing ->

Vendor 03 / 10 · #grafana

Grafana Cloud

Managed observability platform from Grafana Labs with a native CockroachDB integration, prebuilt dashboards, and prebuilt alerts.

Best for

Teams already using Grafana for visualization who want a managed Prometheus-compatible backend
Kubernetes-centric shops scraping CockroachDB with Grafana Alloy alongside cluster metrics
Buyers who want prebuilt CockroachDB dashboards without building their own

Pricing

Usage-based: per-1,000 active series for metrics, per-GB for logs, traces, and profiles
Free tier with limited active series and log volume; paid plans add a platform fee plus usage
The bill grows with active series count, which CockroachDB’s high-cardinality metrics inflate quickly

Pros

Native integration with 4 prebuilt dashboards (Replicas, Runtime, SQL, Storage) and 8 alerts including instance flapping, liveness mismatch, disk low, and unavailable ranges
Scrapes /_status/vars via Grafana Alloy or the Kubernetes Monitoring Helm chart
Enterprise CockroachDB data source plugin lets you query SQL directly from Grafana
Cockroach Labs publishes starter Grafana dashboards for self-hosted users

Cons

Metrics arrive at Prometheus scrape intervals, typically 15 to 60 seconds, not per-second
Active-series pricing penalizes exactly the high-cardinality metrics CockroachDB emits
No SQL-level query analysis; dashboards focus on cluster health

Verdict

Grafana Cloud is the strongest managed option if your team already thinks in Grafana. The native integration is real: four dashboards and eight alerts tuned to CockroachDB failure modes like liveness mismatch and unavailable ranges, not generic up-down checks. The limits are structural. Scrape intervals mean you see a 30-second average of a 5-second problem, and per-series pricing means a busy CockroachDB cluster costs more to observe than a quiet one. Self-hosted Grafana OSS removes the SaaS bill but puts storage and retention back on your plate.

Read full Netdata vs Grafana -> Grafana pricing ->

Vendor 04 / 10 · #prometheus

Prometheus

Open-source time-series database and alerting system, and the canonical way to scrape CockroachDB’s /_status/vars endpoint.

Best for

Teams that want full control of the metrics pipeline and are comfortable operating a time-series database
Kubernetes-native environments where Prometheus is already the standard
Cost-conscious teams that prefer to self-host rather than pay per series or per GB

Pricing

Open source and self-hosted: no license fee, but you operate it yourself
Real operating costs: storage, retention, high availability, and on-call expertise
Managed variants (Grafana Cloud, Amazon Managed Prometheus) shift you back to usage-based pricing

Pros

CockroachDB’s official docs include a step-by-step Prometheus scraping guide
Cockroach Labs publishes starter Grafana dashboards built for Prometheus data
PromQL is the de facto query language for CockroachDB metrics
Alertmanager provides flexible routing, and any Prometheus-compatible tool can consume the same endpoint

Cons

No built-in CockroachDB dashboards: you assemble everything in Grafana yourself
Default 15-second scrape interval misses sub-15-second spikes
Storage, retention, HA, and scaling are your operational problem
No SQL-level query analysis or trace correlation

Verdict

Prometheus is the foundation every other tool on this list stands on: CockroachDB was designed to be scraped by it, and the official documentation walks you through it. But Prometheus is a component, not a monitoring solution. Dashboards, alert content, long-term storage, and HA are assembly required, and the engineers who assemble them are the real cost. If you already run Prometheus well, adding CockroachDB is a scrape config and an afternoon. If you do not, budget for the learning curve before you budget for the hardware.

Read full Netdata vs Prometheus -> Prometheus site ->

Vendor 05 / 10 · #newrelic

New Relic

Usage-based SaaS observability platform with a curated CockroachDB (Prometheus) quickstart containing a dashboard and an alert pack.

Best for

Teams already on New Relic who want CockroachDB metrics beside APM, logs, and infrastructure
Organizations that want prebuilt CockroachDB alerts without writing their own
Kubernetes deployments shipping CRDB metrics via the Prometheus Agent

Pricing

Usage-based: per-GB ingested plus per-user licenses
Free tier with a limited monthly ingest allowance
The bill grows with data volume and user count; high-volume database telemetry is the expensive kind

Pros

CockroachDB quickstart with 1 curated dashboard and 13 alerts covering certificate expiry, node down, slow latch and raft requests, low disk, high FD count, and version mismatch
Ingests via Prometheus remote_write or the Prometheus Agent
Pivots from CRDB metrics into APM, logs, and infrastructure views
Free tier is genuinely useful for evaluation

Cons

SaaS-only: telemetry leaves your infrastructure
Per-GB pricing punishes high-cardinality, high-volume CockroachDB metrics
Quickstart is Prometheus-metric-based; no SQL-level query analysis

Verdict

New Relic’s CockroachDB quickstart is better than most vendors’ first-party integrations: thirteen alerts covering the failure modes that actually page you, including certificate expiry and slow raft requests, which many competitors ignore. The catch is the same as Datadog’s, priced differently: per-GB ingestion means a chatty CockroachDB cluster costs more to watch than a quiet one, and you cannot self-host your way out of it. A solid choice if New Relic is already your observability home; an expensive new one if it is not.

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

A note on pairing tools

The bottom half of this ranking is not worse software; it is software solving a narrower slice of the problem. DBmarlin goes deeper on SQL than anything above it but sees nothing outside the database. Zabbix, SigNoz, Dynatrace, and Instana all watch CockroachDB competently while treating it as one more endpoint in a large estate.

Many production teams end up with two tools: one for infrastructure and cluster health with fast alerting, and one for query-level depth when a slow statement needs a plan, not a chart. If that sounds like your environment, read the remaining cards as complements rather than competitors, and check which combinations duplicate the metrics you already collect.

Vendor 06 / 10 · #dbmarlin

DBmarlin

Database performance monitoring purpose-built for query analysis, wait states, and change tracking, with a native CockroachDB integration via crdb_internal system tables.

Best for

DBAs who need deep SQL-level query analysis for CockroachDB
Teams monitoring CockroachDB alongside Oracle, SQL Server, MySQL, or PostgreSQL in one tool
Organizations that want change tracking and before-after comparison for database changes

Pricing

Per-database-node licensing, not telemetry volume
Free Starter tier with a limited number of licenses
The bill grows with node count, which is predictable but adds up on large clusters

Pros

Native integration via crdb_internal system catalogs, not just Prometheus metrics
Deep SQL performance insight: query bottlenecks, statement plans, transaction contention, and wait states
Time comparison for before-and-after analysis of changes
Listed in CockroachDB’s official third-party monitoring integrations

Cons

Self-hosted CockroachDB only; no CockroachDB Cloud support per CockroachDB’s docs
No infrastructure or host-level monitoring; pair it with another tool for node metrics
Reviewers note UI and performance issues
Per-node licensing grows linearly with cluster size

Verdict

DBmarlin is the only database specialist in this ranking, and it shows: where every other tool reads the metrics endpoint, DBmarlin reads the crdb_internal catalogs and tells you which statement is waiting on what. For query tuning, contention analysis, and change validation, nothing else here comes close. The boundaries are just as clear. It does not monitor hosts, it does not support CockroachDB Cloud, and it will not page you when a node’s disk fills up. Treat it as the second tool in a pair, not the only one.

Visit DBmarlin -> DBmarlin pricing ->

Vendor 07 / 10 · #zabbix

Zabbix

Enterprise-class open-source monitoring platform with an official CockroachDB HTTP template covering the Prometheus and health endpoints.

Best for

Teams that want a full monitoring platform with built-in alerting and no per-host or per-GB fees
Organizations with existing Zabbix deployments adding CockroachDB
Enterprises that prefer open source with optional commercial support

Pricing

Open source and self-hosted: no license fees, you operate it yourself
Optional paid support subscriptions based on response coverage
Zabbix Cloud SaaS available on usage-based tiers

Pros

Official CockroachDB by HTTP template collects from /_status/vars, /health, and /health?ready=1 without external scripts
Bulk collection gathers most metrics in one pass
Triggers for node health, readiness, clock offset, FD limits, SQL errors, certificate expiry, and storage capacity
Per-store low-level discovery for storage, ranges, replicas, and RocksDB metrics

Cons

Template tested against CockroachDB 21.2.8; metric coverage can vary with your CRDB version
No SQL-level query analysis; node and cluster health only
Default polling is minute-level, not per-second
The UI feels dated next to modern observability platforms

Verdict

Zabbix’s official CockroachDB template is better than its reputation suggests: health endpoints, readiness, clock offset, and certificate expiry triggers are exactly the conditions that page you at 3 a.m., and they work without writing a single script. You trade resolution and polish for it. Minute-level polling smooths over short spikes, the template’s version coverage deserves a test against your cluster before you rely on it, and the interface has not kept up. For teams already running Zabbix, enabling the template is close to free value. As a greenfield choice, it asks a lot of patience.

Read full Netdata vs Zabbix -> Zabbix subscriptions ->

Vendor 08 / 10 · #signoz

SigNoz

Open-source, OpenTelemetry-native observability platform ingesting Prometheus and OTel metrics alongside logs and traces.

Best for

Teams that want an open-source Datadog alternative with OTel-native ingestion
Organizations unifying CockroachDB metrics, logs, and application traces in one platform
Kubernetes-centric teams preferring OTel collectors over proprietary agents

Pricing

Cloud: usage-based per-GB for logs and traces, per-million-samples for metrics
Open-source self-hosted option: no license fee, you operate it
The bill grows with data volume either way

Pros

OTel-native ingestion ships CockroachDB Prometheus metrics beside application traces and logs
Open-source core with cloud or self-hosted deployment
Custom dashboards for query latency, node health, and replica behavior
Lower per-GB cost than the major commercial platforms

Cons

No dedicated CockroachDB integration or prebuilt CRDB dashboards; you build them yourself
Lighter prebuilt CockroachDB content than any specialist tool
Self-hosted operation requires ClickHouse expertise

Verdict

SigNoz is a reasonable bet for teams standardizing on OpenTelemetry: scrape CockroachDB’s endpoint with the OTel collector and your database metrics live next to the traces of the applications hitting them. What you do not get is any CockroachDB-specific help. No prebuilt dashboards, no CRDB alert pack, no statement-level analysis; every chart is one you built. If you are already investing in OTel instrumentation that trade can be worth it. If you wanted CockroachDB monitoring out of the box, look higher on this list.

Read full Netdata vs SigNoz -> SigNoz pricing ->

Vendor 09 / 10 · #dynatrace

Dynatrace

AI-powered SaaS observability platform that ingests Prometheus-exposed CockroachDB metrics and correlates them with traces and host signals.

Best for

Enterprises that want AI-assisted anomaly detection via Davis AI on CockroachDB metrics
Teams correlating CockroachDB metrics with application traces and host-level signals
Organizations with existing Dynatrace investments at scale

Pricing

Per-host pricing based on host units (8 GiB per host unit), split between infrastructure and full-stack monitoring
Per-GB pricing for log ingestion
The bill grows with host count and log volume; enterprise contracts often carry minimums

Pros

Ingests Prometheus-exposed CockroachDB metrics from self-hosted or Cloud clusters
Davis AI provides automatic root-cause analysis and anomaly detection
Smartscape topology mapping shows CockroachDB dependencies in context
Geo-aware labeling suits multi-region CockroachDB deployments

Cons

No dedicated CockroachDB sensor or prebuilt dashboard; Prometheus ingestion is configured by hand
Per-host pricing plus log volume gets expensive on large fleets
Steep learning curve and complex configuration

Verdict

Dynatrace can watch CockroachDB, but CockroachDB is not something Dynatrace has thought about specifically. You wire up Prometheus ingestion yourself, and what you get in return is the platform’s real strength: Davis AI correlating a database slowdown with the deployment, host, or network change that caused it, plus topology mapping that places the cluster in your estate automatically. For enterprises already licensed, that correlation justifies the setup effort. For a team buying primarily to monitor CockroachDB, the combination of manual configuration, per-host pricing, and the learning curve is hard to defend.

Read full Netdata vs Dynatrace -> Dynatrace pricing ->

Vendor 10 / 10 · #instana

IBM Instana

IBM’s AI-powered observability platform with a dedicated CockroachDB sensor that automatically discovers clusters and collects metrics.

Best for

Teams that want zero-configuration auto-discovery of CockroachDB clusters
IBM-centric shops that value ecosystem integration
Organizations wanting CockroachDB metrics inside full-stack APM with automatic dependency mapping

Pricing

Per-MVS (Managed Virtual Server) pricing in Essentials and Standard tiers
Pay-per-use option billed per MVS hour
The bill grows with monitored host count; dynamic Kubernetes fleets are hard to forecast under this model

Pros

Dedicated CockroachDB sensor with automatic cluster discovery and metric collection
Dynamic Graph maps CockroachDB dependencies automatically
Collects latency, replica lag, connection counts, and other CRDB metrics
Kubernetes-aware with strong IBM Cloud and hybrid integration

Cons

Per-MVS pricing is complex to forecast for elastic fleets
Known UI usability issues and a steep learning curve
CRDB metric depth is shallower than database specialists like DBmarlin

Verdict

Instana’s dedicated CockroachDB sensor is the easiest start in this list after Netdata: deploy the agent and clusters are discovered and instrumented automatically, with latency, replica lag, and connection metrics flowing without configuration. The depth stops at cluster health, though; there is no statement-level analysis, and the sensor’s metric set is thinner than the integration count suggests. Per-MVS pricing also rewards stable estates and punishes elastic ones. For IBM-aligned enterprises it is a coherent choice. Everyone else should weigh the auto-discovery convenience against the pricing model carefully.

Read full Netdata vs Instana -> Instana pricing ->

Frequently asked questions

What is the best CockroachDB monitoring tool?

Netdata ranks #1 on this list because it is the only tool that pairs a dedicated CockroachDB collector with per-second host metrics, ML anomaly detection on every metric, and per-node pricing that does not penalize high-cardinality telemetry. Datadog has the deepest first-party SaaS integration for teams already on the platform, Grafana Cloud is the strongest managed option for Grafana users, and DBmarlin is the specialist choice for SQL query tuning.

How do I monitor CockroachDB?

Every CockroachDB node exposes a Prometheus-format metrics endpoint at /_status/vars, and CockroachDB Cloud Standard and Advanced tiers expose one as well. The built-in DB Console shows cluster-internal state but is not designed for production alerting. Point an external tool at the endpoint: Netdata, Datadog, Grafana Cloud, and Prometheus all have documented paths, and CockroachDB’s docs note that any Prometheus-compatible tool can consume the metrics.

What are the most important CockroachDB metrics to watch?

Does CockroachDB have built-in monitoring?

Can I monitor CockroachDB with Prometheus and Grafana?

What is the best way to alert on CockroachDB issues?

How much does CockroachDB monitoring cost?

It depends on the pricing model. Per-node (Netdata), per-host (Datadog, Dynatrace, Instana), per-GB (New Relic, SigNoz), per-active-series (Grafana Cloud), and per-database-node (DBmarlin) all scale differently, and high-cardinality CockroachDB telemetry hits per-GB and per-series models hardest. Open-source options like Prometheus, Zabbix, and self-hosted SigNoz avoid license fees but carry real operating costs. The DB Console itself is included with the database.

Can I monitor CockroachDB Cloud with third-party tools?

What is the difference between DB Console and external monitoring tools?

The DB Console is CockroachDB’s built-in web UI for cluster-internal state: current queries, jobs, and node health. External tools add alerting with external delivery, cross-stack correlation with hosts and applications, long-term historical retention, and team access controls. CockroachDB also documents differences in how metrics are computed between third-party integrations and the DB Console, so expect numbers to differ slightly.

How do I monitor CockroachDB SQL query performance?

The DB Console Statements page shows per-statement statistics out of the box. For deeper analysis, DBmarlin reads crdb_internal system catalogs for query bottlenecks, plans, and wait states, and Netdata’s top-queries and running-queries functions pull from crdb_internal.cluster_statement_statistics and SHOW CLUSTER STATEMENTS. For a single misbehaving query, SHOW TRACE FOR SESSION remains the sharpest tool.

The 10 best CockroachDB monitoring tools, ranked by an operator

How we evaluated CockroachDB monitoring tools

Netdata

Datadog

Grafana Cloud

Prometheus

New Relic

DBmarlin

Zabbix

SigNoz

Dynatrace

IBM Instana

Frequently asked questions

Book Your Free Demo