---
title: "What are data sources? | Grafana Labs"
description: "Understanding data sources in Grafana."
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

# A connection to where your data lives

A data source is a **connection** to a backend that holds your data, such as a Prometheus server, a Loki instance, a SQL database, or a cloud monitoring service. Every data source:

- Comes from a data source **plugin**, the connector software. Common ones are built in; for others you install the plugin first.
- Is the **configured instance** you add from that plugin, with its own connection details. You install a plugin, but you configure a data source.
- Comes with a **query editor** that formulates queries in the source’s own language
- Lets you **query, visualize, and alert** on your data without moving it

Your data already lives across systems like these. Grafana connects to them as data sources without custom integration code or a separate data pipeline, querying each one through its API instead of collecting your data with an agent. Your data stays where it lives, with no migration.

[![Grafana sends a query to your data source and the data comes back. Your data stays in its original system.](query-flow.svg)](query-flow.svg "Grafana sends a query to your data source and the data comes back. Your data stays in its original system.")

## What connecting gives you

With that connection, you can:

| Capability    | What it means                                                                                                                                                                              |
|---------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| **Query**     | Run queries against your data from Grafana, using each source’s query language.                                                                                                            |
| **Visualize** | Build dashboards and panels from those queries.                                                                                                                                            |
| **Alert**     | Create alert rules that watch your data and notify you. A single alert rule can query more than one data source, for example a Prometheus error rate alongside an Elasticsearch log query. |

You can also combine multiple data sources in a single dashboard, for example Prometheus metrics alongside MySQL application data and CloudWatch stats.