Grafana VPS Docker Hosting

Grafana is an open-source observability and data visualization platform from Grafana Labs that lets teams query, visualize, alert on, and explore metrics, logs, and traces from many different data sources in one place. It ships dashboards with dozens of panel types, templating variables, unified alerting with routing to Slack, PagerDuty, email, and webhooks, an Explore mode for ad-hoc queries, annotations for marking deploys and incidents, and provisioning as code so dashboards and data sources can be version-controlled across environments.

Grafana itself is lightweight. The official minimum is roughly 512 MB of RAM and a single CPU core for a small instance, since Grafana stores no metrics and pushes the heavy query work down to the data sources it connects to. Real-world deployments comfortably run on 1 to 2 CPU cores and 1 to 2 GB of RAM, with more headroom added as concurrent users, the number of dashboards, and the volume of alert rule evaluations grow.

Grafana listens on TCP port 3000 by default for its web UI and HTTP API. Production deployments place an HTTPS-terminating reverse proxy such as NGINX in front of port 3000, with TLS certificates from Let's Encrypt and the upstream pointing at 127.0.0.1:3000 on loopback. The root_url and serve_from_sub_path settings in grafana.ini are adjusted when Grafana is served under a subpath rather than at the domain root.

A small Grafana instance serving a handful of users runs comfortably on 512 MB to 1 GB of RAM, because Grafana stores no metrics itself and offloads query execution to its data sources. A busier instance with many concurrent viewers, large dashboards, and frequent alert rule evaluations wants 1 to 2 GB of RAM and 1 to 2 CPU cores, and the figure scales mainly with concurrent users and the number of panels rendered at once rather than with data volume.

Backing up Grafana means backing up its configuration database and a few files. A dump of the PostgreSQL or MySQL config database (or a copy of grafana.db on SQLite) preserves dashboards, users, data sources, and alert rules, while the grafana.ini config file and any provisioning YAML capture the rest. Because Grafana queries data sources directly and stores no metrics or logs of its own, the config database plus provisioning files are all that need to be preserved to rebuild an instance from scratch.

Yes. The Grafana open-source edition is fully free and released under the AGPLv3 license, which the core was relicensed to from Apache 2.0 in 2021, for any self-hosted deployment without seat or capacity fees. Grafana Labs also sells Grafana Enterprise, which adds fine-grained RBAC, enterprise data source plugins, reporting, and SAML, and runs the managed Grafana Cloud, which has a free tier and usage-based paid plans built on the same engine.

Grafana stores its own internal state — dashboards, users, organizations, data source definitions, alert rules, and settings — in a configuration database that is separate from the data sources you query. The default is a bundled SQLite file, which is fine for a single node, but PostgreSQL (preferred) or MySQL is the supported choice for production and is required for high-availability deployments running more than one Grafana instance behind a load balancer.

Two install paths are common. Run the official grafana/grafana Docker image mapped to port 3000 on the host, or install the binary or OS package from the Grafana APT/YUM repositories. In either case, point Grafana at a separate PostgreSQL or MySQL configuration database through the GF_DATABASE_* environment variables or the [database] section of grafana.ini before going to production, since moving off the default SQLite file later means migrating the config database under downtime.

Yes, in two distinct roles. Grafana supports PostgreSQL as one of the data sources you query and visualize, alongside Prometheus, Loki, Tempo, Graphite, InfluxDB, Elasticsearch, MySQL, CloudWatch, and many other connectors. PostgreSQL is also the preferred configuration database for storing Grafana's own internal state such as dashboards, users, data source definitions, and alert rules, and it is the recommended backend for any high-availability setup.

LGTM is Grafana Labs' open-source observability stack: Loki for logs, Grafana for visualization and alerting, Tempo for distributed traces, and Mimir for long-term, horizontally scalable Prometheus metrics. Grafana sits at the front as the unified query and dashboard layer over all three signal types, so teams self-hosting the full stack get logs, metrics, and traces correlated in one UI without paying for a hosted observability vendor.