Pick the type of tool before the brand. If a DBA team looks after a SQL Server estate, a dedicated monitor such as Redgate Monitor, SolarWinds SQL Sentry or Idera SQL Diagnostic Manager gives the deepest SQL Server view; SolarWinds DPA covers more engines with wait-time analysis. For PostgreSQL only, pganalyze; for MySQL, PostgreSQL and MongoDB on a budget, the open source Percona PMM. If your company already pays for Datadog, New Relic, Dynatrace or Grafana Cloud, their database modules put query data next to application traces. If you want to run everything yourself for free, Prometheus with exporters and Grafana, Zabbix or pgwatch. On a managed cloud database, switch on the provider's own tool first: CloudWatch Database Insights, Azure SQL database watcher or Cloud SQL Query Insights.
"Database monitoring" covers tools that answer different questions. Dedicated database monitors are built for DBAs: they collect wait statistics, query statistics and execution plans, blocking and deadlocks, and instance health, and keep a history you can go back to after an incident. Observability platforms started with infrastructure or application monitoring (APM) and added database modules, so a slow API call can be traced to the query behind it. Open source stacks give you metrics, dashboards and alerts for no licence fee, at the cost of running them yourself. Cloud-native tools come with a managed database service and are usually the cheapest first step on that platform.
Every one of these tools reads the same built-in sources: SQL Server's DMVs, Query Store and Extended Events; PostgreSQL's pg_stat_statements and pg_stat_* views; MySQL's Performance Schema and sys schema. The products differ in how often they sample, how long they keep data, how they present it and what they alert on. For engine-specific advice see SQL Server monitoring tools, PostgreSQL monitoring tools and MySQL monitoring tools; for the free route, open source database monitoring tools. If you are unsure whether you need database monitoring or APM, read database monitoring vs APM.
Quick picks
Agentless SQL Server monitoring with self-hosted or SaaS editions, and coverage of PostgreSQL, Oracle, MySQL and MongoDB in the same tool.
Agentless wait-time analysis across SQL Server, Oracle, MySQL, MariaDB, PostgreSQL, Db2, SAP HANA and SAP ASE.
PostgreSQL-specific query, plan, log and VACUUM analysis, as SaaS or a self-hosted Enterprise Server.
Open source (server AGPLv3, client Apache 2.0) with Query Analytics for MySQL, PostgreSQL and MongoDB.
Query metrics, samples and explain plans for eight engines, linked to Datadog APM and infrastructure data.
Built into RDS and Aurora, with Standard mode on by default and the replacement for Performance Insights.
How we chose
We considered dedicated monitors (Redgate Monitor, SolarWinds SQL Sentry, SolarWinds Database Performance Analyzer, Idera SQL Diagnostic Manager, Percona Monitoring and Management, pganalyze, Quest Spotlight on SQL Server Enterprise), observability platforms with database features (Datadog Database Monitoring, New Relic, Dynatrace, Grafana Cloud Database Observability, SolarWinds Observability SaaS), open source stacks (Prometheus with exporters and Grafana, Zabbix, pgwatch, Netdata, pgBadger) and the three large clouds' own tools. To be listed, a product had to be sold or actively developed in October 2026, document which engines it supports, and state its deployment model and pricing basis or licence on an official page.
Quest Spotlight on SQL Server Enterprise is still supported by Quest, but we could not read enough of its current product pages to verify its details, so it is not an entry. Netdata and pgBadger are covered in the open source guide. SolarWinds Observability SaaS is mentioned under SolarWinds DPA.
For each product we recorded its type, engines, deployment (agent or agentless, SaaS or self-hosted), what it collects, and how it is priced. Where a price is quoted it is one example from the vendor's own pricing page, dated; we do not estimate anyone's bill. The order is not a ranking: entries are grouped as dedicated monitors, observability platforms, open source stacks and cloud-native tools.
At a glance
| Tool | Type | Engines | Deployment | Pricing model | Main trade-off |
|---|---|---|---|---|---|
| Redgate Monitor | Dedicated monitor | SQL Server, PostgreSQL, Oracle, MySQL, MongoDB | Self-hosted (agentless) or SaaS | Per server licence, subscription | No public prices |
| SolarWinds SQL Sentry | Dedicated monitor | SQL Server, Azure SQL, SSAS | Self-hosted, agentless; Windows service plus SQL Server repository | Per monitored target, subscription | SQL Server family only |
| SolarWinds DPA | Dedicated monitor (wait-time analysis) | SQL Server, Oracle, MySQL, MariaDB, PostgreSQL, Db2, SAP HANA, SAP ASE | Self-hosted, agentless | Per instance, subscription | Less instance-health detail than SQL Server specialists (editorial) |
| Idera SQL Diagnostic Manager | Dedicated monitor | SQL Server, Azure SQL, RDS for SQL Server | Self-hosted, agentless | Per instance, subscription | SQL Server only |
| Percona PMM | Open source monitor | MySQL, MariaDB, PostgreSQL, MongoDB, Valkey, Redis | Self-hosted server plus client agents | Free; paid Percona support | You run and upgrade the server |
| pganalyze | Dedicated monitor | PostgreSQL only | SaaS or self-hosted Enterprise Server; collector agent | Per database server, monthly | PostgreSQL only |
| Datadog Database Monitoring | Observability platform module | Postgres, MySQL, MariaDB, Oracle, SQL Server, MongoDB, DocumentDB, ClickHouse | SaaS; Datadog Agent | Per database host per month | Cost adds up per host on top of other Datadog products |
| New Relic | Observability platform | Query monitoring for PostgreSQL, MySQL, SQL Server | SaaS; infrastructure agent and integrations | GB ingested plus paid users | Usage pricing needs ingest control |
| Dynatrace | Observability platform | SQL Server, Oracle, PostgreSQL, MySQL, MariaDB, SAP HANA and others | SaaS; extensions via ActiveGate | Annual commit consumed by usage (DPS) | Minimum annual commitment |
| Grafana Cloud Database Observability | Observability platform module | MySQL, PostgreSQL | SaaS; Grafana Alloy collector | Per host-hour on Pro | Two engines only |
| Prometheus, exporters and Grafana | Open source stack | Any engine with an exporter | Self-hosted | Free | Metrics only; no query plans; you build dashboards and alerts |
| Zabbix | Open source monitoring platform | SQL Server, MySQL, PostgreSQL, Oracle and more via templates | Self-hosted server, frontend and agents | Free (AGPLv3); paid support | General infrastructure tool; shallow query analysis |
| pgwatch | Open source monitor | PostgreSQL only | Self-hosted (Docker) | Free (BSD-3-Clause) | PostgreSQL only; you run the stack |
| CloudWatch Database Insights | Cloud-native | Aurora, RDS engines, self-managed PostgreSQL on EC2 | AWS managed | Standard free to 7 days; Advanced paid | AWS only |
| Azure SQL database watcher | Cloud-native (preview) | Azure SQL Database, Azure SQL Managed Instance | Azure managed; data in Azure Data Explorer or Fabric | Watcher free; pay for data store | Preview, no SLA |
| Cloud SQL Query Insights | Cloud-native | Cloud SQL for MySQL, PostgreSQL, SQL Server | Google Cloud managed | No extra charge | Google Cloud only; 7 days on Enterprise edition |
Redgate Monitor
Redgate Monitor (14.34.0 when checked), formerly SQL Monitor, now lists support for SQL Server, PostgreSQL, Oracle, MySQL and MongoDB (Oracle, MySQL and MongoDB were added in 14.0.52, March 2025), on premises, in the cloud or hybrid. It is sold in Standard and Enterprise editions, each either self-hosted or as a SaaS edition that Redgate runs for you. Redgate's documentation describes the self-hosted design as agentless: a Base Monitor service connects over WMI (Windows) or SSH (Linux) and T-SQL, and stores data in a SQL Server or PostgreSQL with TimescaleDB repository, with a web server for the console.
Besides performance metrics and query analysis it collects estate data such as configuration, versions and patches, disk use and backups, and integrates alerts with Splunk, ServiceNow and webhooks. Licensing is per server, with tiered discounts; Redgate states that one licence covers five Azure SQL databases. Prices are not shown publicly. See Redgate Monitor vs SQL Sentry and Datadog vs Redgate Monitor.
- No list prices on the public pricing page; you need a quote.
- Self-hosted needs a Windows or Linux host for the Base Monitor plus a repository database you maintain.
- Its depth is greatest on SQL Server; check the features you need for other engines in a trial (editorial).
SolarWinds SQL Sentry
SQL Sentry came to SolarWinds with the SentryOne acquisition and is sold as SolarWinds SQL Sentry. It monitors SQL Server (the current 2026.3 release supports SQL Server 2017 to 2025, including SQL Server on Amazon RDS and EC2), Azure SQL Database, Azure SQL Managed Instance and SQL Server Analysis Services. SolarWinds lists deadlock analysis, blocking, TempDB, query and index analysis, an event calendar of jobs and storage forecasting.
SolarWinds' documentation describes it as agentless: one or more monitoring services poll the targets and write to a SQL Server repository database, viewed through a Windows client or an optional web portal. It is licensed per monitored target, and from version 2024.4 uses the same Database Self Hosted licence type as DPA, so one licence key can be split between the two products. SolarWinds itself was taken private by Turn/River Capital in April 2025. See SQL Sentry vs SolarWinds DPA.
- SQL Server family only.
- Self-hosted infrastructure: a Windows monitoring service and a SQL Server repository to size and maintain.
- No public prices.
SolarWinds Database Performance Analyzer
Database Performance Analyzer (DPA) is built around wait-time analysis: it shows which queries spend time waiting and on what. SolarWinds lists support for Oracle, SQL Server (including Azure SQL), MySQL, MariaDB, PostgreSQL, SAP HANA, IBM Db2 and SAP ASE, including databases on Amazon RDS and Google Cloud SQL, with an agentless design, anomaly detection, tuning advisors and VMware correlation.
DPA (2026.2 when checked) is self-hosted. Current subscription licences are Database Self-Hosted (DBSH) and a DBaaS variant for services such as Azure SQL Database; licences float between instances. SolarWinds also sells Database Observability inside its SaaS platform, SolarWinds Observability; as listed on its pricing page on 7 October 2026, that is USD 70 per database instance per month, billed on a daily high-water mark of active instances. See Redgate Monitor vs SolarWinds DPA, Datadog vs SolarWinds DPA and SolarWinds DPA alternatives.
- No public prices for the self-hosted product.
- Focused on waits and queries rather than full estate inventory (editorial).
- Two overlapping SolarWinds database lines (DPA, SQL Sentry, plus the SaaS platform) make the choice within SolarWinds less obvious.
Idera SQL Diagnostic Manager
Idera describes SQL Diagnostic Manager as agentless SQL Server monitoring for SQL Server on premises, Azure SQL Database and Managed Instance, Amazon RDS for SQL Server, Google Cloud and VMware or Hyper-V hosts. It lists adaptive baselines for alerting, wait statistics, deadlock visualisation, live sessions, forecasting and query tuning features.
Pricing is per instance on subscription, and Idera stresses that there are no per-gigabyte ingest fees and that the customer owns the monitoring repository. It is a direct alternative to SQL Sentry and Redgate Monitor for SQL Server estates; see SQL Sentry alternatives.
- SQL Server only.
- No public prices.
- Self-hosted, so you maintain the repository and console servers.
Percona Monitoring and Management
Percona Monitoring and Management (PMM) is an open source monitoring tool for MySQL (including Percona Server, MariaDB and Amazon RDS and Aurora MySQL), PostgreSQL and MongoDB, and its repository also lists Valkey and Redis. PMM Server is licensed AGPLv3 and PMM Client Apache 2.0. Version 3 is current; the latest release when checked was 3.9.1 (19 August 2026).
PMM Server bundles Grafana dashboards, VictoriaMetrics for metrics, ClickHouse for Query Analytics and PostgreSQL, while PMM Client runs on or near each database host. It adds Percona Advisors (configuration and security checks), alerting and backup management for MySQL and MongoDB. See open source database monitoring tools.
- No SQL Server or Oracle support.
- You host, secure and upgrade PMM Server yourself.
- Commercial support comes from Percona as a separate contract.
pganalyze
pganalyze is a PostgreSQL-only monitoring service. Its collector reads pg_stat_statements, EXPLAIN plans for slow queries, PostgreSQL logs and schema statistics, and supports Amazon RDS and Aurora, Azure Flexible Server, Google Cloud SQL and AlloyDB, Heroku, Crunchy Bridge, Aiven, Supabase, Neon and self-managed servers. The data goes to pganalyze's cloud or to a self-hosted Enterprise Server.
As listed on its pricing page on 7 October 2026, the Production plan is USD 149 a month for one database server with 14 days of history, Scale is USD 399 a month for up to four servers with 35 days, and Enterprise is custom priced with 100 days and self-hosting. Replicas count as half a server. See PostgreSQL monitoring tools.
- PostgreSQL only.
- Retention is tied to the plan: 14 days on Production.
- Self-hosting requires the custom-priced Enterprise plan.
Datadog Database Monitoring
Datadog documents Database Monitoring for self-hosted and managed versions of Postgres, MySQL, MariaDB, Oracle, SQL Server, MongoDB, Amazon DocumentDB and ClickHouse. The Datadog Agent collects normalised query metrics, individual query samples, explain plans and host metrics, and the data sits alongside Datadog APM and infrastructure monitoring.
As listed on Datadog's pricing list on 7 October 2026, Database Monitoring is USD 70 per database host per month billed annually, or USD 84 on demand. See Datadog vs New Relic, Prometheus vs Datadog and Datadog vs SolarWinds DPA.
- SaaS only; monitoring data leaves your network.
- Per-host pricing is added to any other Datadog products you use, so total cost needs modelling.
- Fewer DBA-specific estate features (backups, jobs, licensing inventory) than dedicated monitors (editorial).
New Relic
New Relic's query performance monitoring, generally available since July 2025, collects slow queries, wait events, execution plans and blocking sessions from PostgreSQL, MySQL and Microsoft SQL Server through its infrastructure integrations, without needing APM agents.
Pricing is by data ingested and by user type rather than by host. As listed on New Relic's pricing page on 7 October 2026, the free tier includes 100 GB of ingest a month, one full platform user and unlimited basic users, and ingest beyond that costs USD 0.40 per GB on the original option. See Datadog vs New Relic and database monitoring vs APM.
- Query-level monitoring covers three engines.
- Cost depends on ingest volume and full-user count, which need controls.
- SaaS only.
Dynatrace
Dynatrace monitors databases through extensions that run on an ActiveGate and connect remotely (its documentation covers PostgreSQL, MySQL, SQL Server, Oracle, MariaDB and SAP HANA among others), plus host metrics from OneAgent. It collects SQL statements and execution plans, sessions, locks and deadlocks, and stores the data in its Grail data lakehouse.
Dynatrace sells a Dynatrace Platform Subscription: a minimum annual spend commitment consumed against a rate card. As listed on its pricing page on 7 October 2026, Infrastructure Monitoring is USD 0.04 per host-hour; there is a 15-day trial.
- Requires an annual commitment.
- Database monitoring is one part of a large platform, which is a lot to adopt only for databases (editorial).
- Remote database extensions need ActiveGate infrastructure.
Grafana Cloud Database Observability
Grafana Cloud Database Observability surfaces query performance data, explain plans, query samples, wait events and table schemas from MySQL and PostgreSQL, self-managed or on AWS, Azure and Google Cloud. Grafana Alloy collects the telemetry and sends it to Grafana Cloud.
As listed on Grafana's pricing page on 7 October 2026, Database Observability costs USD 0.07 per host-hour on the Pro plan, which has a USD 19 monthly platform fee. The free Grafana Cloud tier (10,000 active metric series, 50 GB of logs, 14 days' retention) suits Prometheus-style database metrics. See Prometheus vs Datadog.
- Query-level features cover MySQL and PostgreSQL only.
- Usage-based pricing across series, logs and hosts needs watching.
- SaaS; the self-hosted route is the open source stack, which lacks this module.
Prometheus, exporters and Grafana
Prometheus (Apache 2.0) scrapes metrics from exporters: postgres_exporter (prometheus-community) and mysqld_exporter (Prometheus project), both Apache 2.0, and for SQL Server the MIT-licensed sql_exporter, which Grafana Alloy also embeds as its prometheus.exporter.mssql component. Grafana (AGPLv3) draws the dashboards, and Prometheus Alertmanager handles alerts.
The result is a consistent metrics platform across engines and the rest of your infrastructure. It is metric-based: you get counters and gauges from the engine's statistics views, not individual query samples or execution plans. Prometheus keeps 15 days of local data by default and its local storage is not clustered, so long-term retention needs remote storage. See Prometheus vs Datadog and open source database monitoring tools.
- No query plans or per-execution samples.
- You assemble and maintain exporters, dashboards, alert rules and storage.
- Long retention needs extra components.
Zabbix
Zabbix is an open source monitoring platform, licensed AGPLv3 from version 7.0 (earlier versions were GPLv2); 7.4 is the current release. Official templates cover SQL Server (through Zabbix agent 2 or ODBC), MySQL, PostgreSQL and other databases, collecting performance counters, database state and size, backups, SQL Agent jobs and availability group state.
Zabbix needs a server, a PHP web frontend and a backend database (MySQL, MariaDB, PostgreSQL or TimescaleDB). It is strongest as estate-wide availability and threshold alerting rather than query tuning.
- Little query-level analysis compared with dedicated monitors (editorial).
- Several components to run and tune.
- Templates need adjustment to avoid noisy alerts (editorial).
pgwatch
pgwatch, from CYBERTEC, is a PostgreSQL-specific monitoring tool under the BSD-3-Clause licence, currently at major version 5. It runs a collector with a web UI, stores metrics in PostgreSQL or TimescaleDB (or exposes them to Prometheus), and ships pre-built Grafana dashboards.
- PostgreSQL only.
- You run the collector, metrics database and Grafana.
Amazon CloudWatch Database Insights
CloudWatch Database Insights is AWS's fleet monitoring for Aurora MySQL and PostgreSQL, RDS for SQL Server, MySQL, PostgreSQL, Oracle and MariaDB, and self-managed PostgreSQL on EC2. It shows database load broken down by waits, SQL, hosts and users. AWS states that Performance Insights reached end of life on 31 July 2026 and its users were moved to Database Insights.
Standard mode is on by default and includes 7 days of detailed metrics free, with paid retention of 1 to 24 months. Advanced mode adds per-query statistics, slow query analysis, execution plans (Aurora PostgreSQL, RDS for Oracle and RDS for SQL Server) and fleet views, priced on the CloudWatch pricing page. See AWS RDS vs Azure SQL Database.
- AWS databases only, plus self-managed PostgreSQL on EC2.
- Execution plans and slow query analysis need Advanced mode.
- Some Advanced mode features vary by region.
Azure SQL database watcher
Database watcher is Microsoft's managed monitoring for Azure SQL Database and Azure SQL Managed Instance, and was still in preview when checked. It collects data from more than 70 catalog views and DMVs (active sessions, top queries, waits, index metadata, SQL Agent jobs, backups) into Azure Data Explorer or Real-Time Analytics in Microsoft Fabric, where you can query it with KQL or T-SQL, with estate and resource dashboards and Azure Monitor alerts.
Microsoft charges nothing for the watcher or dashboards and nothing per monitored database; you pay for the data store, alerts and any cross-region bandwidth, and a free Azure Data Explorer cluster can be used where no SLA is needed. A watcher can monitor up to 100 targets. See Azure SQL Database vs SQL Server.
- Preview: no SLA, and limits may change.
- Azure SQL only; it does not monitor SQL Server on your own servers.
- Alerts are not available with a free Azure Data Explorer cluster or with Fabric as the store.
Google Cloud SQL Query Insights
Query Insights is built into Cloud SQL for MySQL, PostgreSQL and SQL Server, and Google states there is no additional cost for it. It tracks query load and can tag queries by application model, view, controller, route, user and host. On Enterprise edition it keeps 7 days of metrics and samples up to 20 query plans a minute; Enterprise Plus extends this to 30 days and 200 plans a minute, and adds wait event analysis and an index advisor (PostgreSQL figures).
- Cloud SQL only.
- 7-day retention on Enterprise edition.
- Wait events and index advice need Enterprise Plus.
How to choose
Start with the built-in tools, because every product above reads them: Query Store and DMVs for SQL Server, pg_stat_statements for PostgreSQL, Performance Schema for MySQL. On a managed cloud database, turn on the provider's tool before buying anything. Buy a dedicated monitor when a DBA team needs history, alerting and estate views across many instances, especially for SQL Server. Choose an observability platform module when developers and SREs need query data next to application traces, and the platform is already paid for. Choose an open source stack when you have the people to run it and want no licence fee or full control of the data.
Then compare on four practical points: which engines and hosting models are covered, what is collected (waits, query samples, plans, OS counters) and for how long, how alerts are configured and routed, and the pricing basis (per server, per host, per host-hour or per GB). Run a trial against your own busiest instance, and ask the vendor how its collection affects the monitored server rather than relying on a general claim.
Frequently asked questions
What is the difference between database monitoring and APM?
APM follows requests through application code and shows which calls are slow; database monitoring looks inside the engine at waits, query statistics, plans, locks and resources. APM can show that a query is slow; database monitoring explains why. Many teams use both. See database monitoring vs APM.
Is there a free database monitoring tool?
Yes. Percona PMM, Prometheus with exporters and Grafana, Zabbix and pgwatch are free and open source, and the cloud providers' own tools have free levels (CloudWatch Database Insights Standard mode, Cloud SQL Query Insights, and the database watcher itself, which only charges for its data store). See open source database monitoring tools.
What replaced Amazon RDS Performance Insights?
CloudWatch Database Insights. AWS states that Performance Insights reached end of life on 31 July 2026 and that users were moved to Database Insights, where Standard mode keeps flexible retention at the same price and Advanced mode adds execution plans and on-demand analysis.
Do monitoring tools slow the database down?
Any collection uses some resources. Vendors describe their approaches (agentless polling, sampling intervals, reading from statistics views), but the impact depends on your workload and settings, so test a tool on a busy non-production instance first and watch the monitoring login's own activity.
Which tool is best for SQL Server specifically?
Dedicated SQL Server monitors such as Redgate Monitor, SolarWinds SQL Sentry and Idera SQL Diagnostic Manager, built on the free baseline of Query Store, Extended Events and DMVs. See SQL Server monitoring tools.
Is SolarWinds still a public company?
No. Turn/River Capital completed its acquisition of SolarWinds on 16 April 2025, and SolarWinds' shares stopped trading. It continues to operate under the SolarWinds name and sells SQL Sentry and DPA.
Sources
- Redgate Monitor product page
- Redgate Monitor pricing
- How Redgate Monitor works (documentation)
- SolarWinds SQL Sentry
- SQL Sentry components and architecture
- SolarWinds Database Performance Analyzer
- DPA licence types
- SolarWinds Observability SaaS pricing
- Turn/River completes acquisition of SolarWinds
- Idera SQL Diagnostic Manager
- Percona Monitoring and Management
- PMM repository and licences
- pganalyze pricing
- pganalyze documentation
- Datadog Database Monitoring documentation
- Datadog pricing list
- New Relic query performance monitoring
- New Relic pricing
- Dynatrace database monitoring documentation
- Dynatrace pricing
- Grafana Cloud Database Observability
- Grafana Cloud pricing
- Prometheus storage documentation
- Zabbix licence
- pgwatch repository
- CloudWatch Database Insights
- Azure SQL database watcher overview
- Cloud SQL Query Insights
Checked October 2026.
How we research these guides: our editorial method.