What is Grafana?
Grafana is an observability and data visualization server that queries metrics, logs and traces where they are already stored and presents them in dashboards, in the ad hoc Explore view and in alert rules. It does not require data to be copied into Grafana first: each panel query is sent to a configured data source such as Prometheus, Loki, Elasticsearch, InfluxDB or PostgreSQL, and a single dashboard can mix several sources. According to the Grafana Labs about page, the dashboarding tool was created as an open source project in December 2013. Grafana Labs also maintains Loki, Mimir, Tempo, k6, Pyroscope, Faro, Beyla and Alloy.
The server is written in Go with a TypeScript and React frontend and is developed in the grafana/grafana repository. Configuration, users and dashboards are stored in a database: SQLite 3 is embedded by default, while MySQL 8.0 or later and PostgreSQL 12 or later are supported and are required for high availability. The installation guide lists packages for Debian and Ubuntu, RHEL and Fedora, SUSE and openSUSE, macOS and Windows, plus Docker images, Kubernetes manifests and a Helm chart. The documented minimum is 512 MB of memory and one CPU core, which the docs describe as a floor for evaluation rather than a production size.
Two self-managed editions are distributed. The Open Source edition is published as the grafana/grafana image and as packages; the Enterprise edition, grafana/grafana-enterprise, is the default download, is free to run and contains all OSS features, and unlocks additional features such as SAML, team sync, reporting and Enterprise data source plugins when an Enterprise license is applied. Grafana Labs also operates Grafana Cloud, a hosted service with a Free tier, a Pro tier and an Enterprise tier described on the pricing page.
The latest release is Grafana 13.3.0, published on October 27, 2026. Its notes list ten CVE fixes and changes across alerting (an import wizard for Mimir Alertmanager configuration and external Alertmanager auto-sync), dashboards (auto grid as the default layout, a $__namespace variable, faster Recently Deleted search), Git Sync provisioning and the home page. The 13.0 line made Git Sync, dynamic dashboards, dashboard restore and the Grafana Advisor health checks generally available, removed the Image Renderer plugin support and replaced the grafana-cli and grafana-server binaries with grafana cli and grafana server, as described in What's new in Grafana 13.0.
Grafana is distributed under AGPL-3.0-only. The repository's LICENSING.md lists directories that use Apache-2.0 instead, including the grafana-data, grafana-ui and grafana-runtime packages used by plugin developers. Grafana itself does not store time series, logs or traces; those stay in the connected backends. Similar projects in this directory include Prometheus, SigNoz and Datadog.
Key capabilities
- Dashboards with template variables Dashboards are made of panels, each with its own queries, transformations and visualization settings. Template variables appear as drop-downs at the top of a dashboard, so one dashboard can be switched between hosts, clusters or services. Since Grafana 13.0 the dynamic dashboards layout engine is on by default, and 13.3 makes the auto grid the default layout.
- Visualization panels Built-in visualizations include time series, bar chart, stat, gauge, table, heatmap, logs, traces, node graph, geomap, canvas and others, documented in the visualizations reference. Panel plugins from the plugin catalog add further types. Field overrides and thresholds control color, units and display per series.
- Data sources queried in place Core data sources include Prometheus, Graphite, InfluxDB, OpenTSDB, Loki, Elasticsearch, Tempo, Jaeger, Zipkin, Pyroscope, MySQL, PostgreSQL, Microsoft SQL Server, AWS CloudWatch, Azure Monitor and Google Cloud Monitoring, as listed in the data sources documentation. Each panel query can target a different source, including in one graph. Grafana 13.0 added a raw query editor with Query DSL and ES|QL for Elasticsearch and an IBM DB2 data source in public preview.
- Explore Explore is a query workspace outside dashboards for ad hoc investigation. It supports split view to compare two queries, time ranges or data sources side by side, query history and an inspector. When moving from metrics to logs, label filters are carried over, and logs can be searched or streamed live.
- Logs, traces and profiles views The logs visualization shows log lines with level coloring, a details panel with fields and actions, and deduplication options. Trace views for Tempo, Jaeger and Zipkin show spans, a critical path and a node graph of service calls. Data links connect a log line or span to related metrics or traces.
- Grafana Alerting Alert rules are defined against data source queries and evaluated on a schedule; multi-dimensional rules create one alert instance per label set. Notifications are routed through contact points and notification policies to Slack, PagerDuty, email, webhooks and other integrations. The alerting documentation also covers silences, notification templates and contact point configuration; release 13.3 extends the import wizard to Mimir Alertmanager configuration files.
- Git Sync and dashboards as code Git Sync, generally available since 13.0, connects Grafana to a Git repository so dashboards and folders can be committed from the UI and synchronized from the repository. Release 13.3 adds generic Git OAuth connections, pull request previews and folder README rendering. Dashboards can also be provisioned from files or managed through the HTTP API.
- Restore deleted dashboards A Recently deleted view lets administrators recover deleted dashboards. Grafana 13.3 moved its search to the server side and hides links to folders that no longer exist.
- Plugins The plugin catalog offers data source, panel and app plugins that can be installed from the UI or preinstalled with the
GF_PLUGINS_PREINSTALLenvironment variable. Plugin developers build on the Apache-2.0 licensed@grafana/data,@grafana/uiand@grafana/runtimepackages. - Authentication and access control Grafana supports basic auth, anonymous access, LDAP, Auth Proxy, JWT Proxy and OAuth with GitHub, GitLab, Google, Okta, Entra ID, Grafana.com or a generic OAuth provider, as listed in the authentication guide. SAML is available in Enterprise and Grafana Cloud only; team sync and enhanced LDAP synchronization are also Enterprise features.
- HTTP API Dashboards, folders, data sources, alert rules, users and other resources are managed through the HTTP API with service account tokens. Grafana 13.0 deprecates the
/apipath in favor of the Kubernetes-style/apisendpoints and disables the numeric-ID data source APIs by default. - Grafana Advisor Introduced as generally available in 13.0, Grafana Advisor runs weekly checks that flag failing data sources, outdated plugins and misconfigured SSO settings in an instance.
- Linux, Windows and macOS packages The download page has tabs for Linux, Windows, Mac, Docker and Linux on ARM64, with DEB packages for Ubuntu and Debian, RPM packages for RHEL, Fedora and SUSE, and standalone tar.gz binaries; on November 10, 2026 it listed 13.3.1 as the latest Enterprise build. The GitHub release assets for 13.3.0 include DEB and RPM packages for amd64, arm64 and ARMv7, Raspberry Pi ARMv6 packages and an SPDX software bill of materials.
- Docker and Kubernetes The documented quick start is
docker run -d -p 3000:3000 --name=grafana grafana/grafana-enterprise, with a volume on/var/lib/grafanafor persistence. Kubernetes manifests and the Grafana Helm chart are documented for cluster installs. High availability requires an external MySQL or PostgreSQL database. - Grafana Enterprise features With an Enterprise license, the Enterprise edition adds role-based access control, data source permissions, query and resource caching, scheduled PDF reports, custom branding, usage insights, auditing, Vault integration and Enterprise data source plugins for Splunk, Snowflake, Datadog, ServiceNow, Jira, MongoDB and others.
- Grafana Cloud The hosted service bundles Grafana with managed metrics, logs, traces and profiles backends. The Free tier includes 10k active metric series, 50 GB each of logs, traces and profiles and 3 active visualization users per month with 14-day retention. Pro starts at 19 USD per month plus usage, and Enterprise starts at a 25,000 USD annual commitment.
- Public demo Grafana Labs runs play.grafana.org, a public instance with example dashboards for many data sources and visualizations.
Tested by iThub
· Version 13.3.0
- Environment
- Windows AMD64; Docker Desktop 4.95.0, Docker Engine 29.8.2, Linux containers; synthetic local data; official Ubuntu image; 512 MiB container limit
- Scenario
- Provision TestData and render a time-series panel from six synthetic values
- Observed result
- The health endpoint reported database OK, TestData reported Data source is working, and the dashboard query returned 1, 20, 90, 30, 5 and 0. The actual Grafana panel rendered those values.
The initial Alpine image failed locally with exec format error after an interrupted pull. The official 13.3.0 Ubuntu image started successfully. The earlier failure was not diagnosed as a Grafana defect.

Scope and limits. Only the built-in synthetic TestData source was exercised, using a local anonymous Viewer. Real databases, authentication policy, alerts, plugins, load and production operations were not tested.
Download the test record (JSON)How we record testsProject source ↗Project source ↗
Where Grafana fits
Infrastructure and Kubernetes dashboards
Operators connect a Prometheus or Mimir data source and build or import dashboards for hosts, containers and clusters. Template variables switch the view between nodes, namespaces or workloads, and stat, gauge and time series panels show CPU, memory, disk and network usage. Prebuilt dashboards are shared on grafana.com and can be imported by ID or JSON.
Correlating metrics, logs and traces
In Explore, a team can query Prometheus metrics next to Loki logs in split view and keep label filters when switching between them. Data links and trace-to-logs settings connect a span in Tempo, Jaeger or Zipkin to the log lines and metrics of the same service. The trace view adds a span timeline, critical path and node graph.
Alerting and on-call notification
Alert rules evaluate queries against any supported data source on a schedule. Multi-dimensional rules produce a separate alert instance per label combination, for example per CPU or per host. Contact points and notification policies route alerts to Slack, PagerDuty, email or webhooks, and silences suppress notifications for matching alerts.
Business and SQL data visualization
Grafana is not limited to time series databases. MySQL, PostgreSQL and Microsoft SQL Server data sources run SQL queries whose results feed tables, bar charts, geomaps and stat panels. Enterprise plugins extend this to sources such as Snowflake, Databricks, Salesforce and Jira.
Dashboards as code
Git Sync stores dashboards and folders in a Git repository, with commits made from the Grafana UI and pull request previews of changes. Alternatively, dashboards, data sources and alert rules can be provisioned from YAML and JSON files at startup or managed through the HTTP API.
Choosing between self-managed and hosted
The same Grafana application is available as packages and containers to run on one's own servers, with SQLite for single instances or MySQL and PostgreSQL for high availability, and as Grafana Cloud, where Grafana Labs operates Grafana together with managed Mimir, Loki, Tempo and Pyroscope backends. Grafana Cloud plans and included usage are listed on the pricing page.
Install
docker run -d -p 3000:3000 --name=grafana grafana/grafana-enterpriseReviews of Grafana
Be the first to share your experience with Grafana.
Write a review
Loading review form…
Found Grafana useful? Upvote it todayYour vote helps other IT teams find it in the daily ranking.






