Common data sources on Grafana Labs

Browse the data sources

Tue, 18 Aug 2026 05:43:38 -0700

Browse the data sources

Grafana supports more than 150 data sources, but you don’t need to learn them all. This journey covers some common ones.

Here are the data sources you can explore in this journey.

Data source	Type	Best for	Query language
Prometheus	Metrics	Cloud-native metrics	PromQL
InfluxDB	Metrics	IoT, high-volume	Flux, InfluxQL, SQL (v3+)
Loki	Logs	Cost-effective logs	LogQL
Elasticsearch	Logs	Full-text search	Lucene
CloudWatch	Cloud	AWS resources	Metrics Insights, metric math
MySQL	SQL	Application data	SQL
PostgreSQL	SQL	Application data	SQL
Infinity	Flexible	APIs, files	JSONata, JQ, UQL
Google Sheets	Flexible	Spreadsheet data	Cell references

A Grafana Cloud account is helpful for hands-on practice, but optional. You can read through this journey to understand the setup process without connecting anything yourself.

Prometheus

Tue, 18 Aug 2026 05:43:38 -0700

Prometheus: Cloud-native metrics

What it’s for: Time-series metrics including CPU, memory, request rates, error counts, and latency.

Good to know

Query language: PromQL
Connects to any Prometheus-compatible backend, including Prometheus, Grafana Mimir, and Thanos
Native Grafana support, so there’s nothing extra to install

Best for

Kubernetes environments
Microservices architectures
Cloud-native applications

Set it up

Open this path in your Grafana Cloud stack for a fully interactive experience, or read through it to understand the process first.

Learning path

Connect to a Prometheus data source

Welcome to the Grafana learning path that provides the best practice for configuring a Prometheus data source plugin.

10 min

Beginner

Docs & blog posts

Open in Grafana Cloud

Complete this learning path directly in your Grafana Cloud stack with an interactive learning experience.

Open in Grafana Cloud

View on website

InfluxDB

Tue, 18 Aug 2026 05:43:38 -0700

InfluxDB: High-throughput time-series database

What it’s for: High-volume metrics including IoT sensors, industrial monitoring, and sub-second precision data.

Good to know

Built for high write throughput, which suits high-volume, high-frequency data
Built-in downsampling and retention policies help manage storage as data ages.
The query language depends on your version.

Query language support by version

InfluxDB version	Query languages
1.x	InfluxQL (Flux in 1.8+)
2.x	Flux, InfluxQL
3.x (including Cloud Serverless)	SQL, InfluxQL

Best for

IoT and sensor data
Industrial monitoring
High-frequency metrics (sub-second)

Documentation

View the full documentation.

InfluxDB data source

Loki

Tue, 18 Aug 2026 05:43:38 -0700

Loki: Cost-effective log aggregation

What it’s for: Logs including application logs, system logs, and audit trails.

Good to know

Query language: LogQL, which is similar to PromQL
Loki indexes labels rather than full log content, which lowers storage cost. You filter by labels first, then search within that subset, instead of running arbitrary full-text search.
It shares Prometheus’s label model, so your Prometheus labels carry over.

Best for

Teams already using Prometheus
Cost-conscious log aggregation
Correlating logs with metrics

Documentation

View the full documentation.

Loki data source

Elasticsearch

Tue, 18 Aug 2026 05:43:38 -0700

Elasticsearch: Full-text search and log analytics

What it’s for: Logs with complex search needs including security analytics, audit logging, and full-text search.

Good to know

Query language: Lucene query syntax
Indexes the full content of every log, which enables full-text search across everything
That full indexing uses more storage, memory, and CPU than label-based stores like Loki, and the cluster takes more work to size and maintain.

Best for

Complex log search requirements
Security analytics (security information and event management, or SIEM)
Existing Elasticsearch installations

Documentation

View the full documentation.

Elasticsearch data source

CloudWatch

Tue, 18 Aug 2026 05:43:38 -0700

CloudWatch: Native AWS monitoring

What it’s for: AWS resources including EC2, RDS, Lambda, S3, ECS, and many other AWS services.

Good to know

Connects to AWS services like EC2, RDS, Lambda, S3, and ECS, with no collection agents to deploy
Requires AWS authentication configured in Grafana
AWS bills the GetMetricData API that Grafana uses per metric retrieved, so dashboards that pull many metrics on frequent refreshes can add cost.

Best for

AWS infrastructure monitoring
Services only available via CloudWatch
Quick visibility into AWS resources

Cost Tip: Monitor your CloudWatch costs. GetMetricData is billed per metric retrieved, so more metrics and more frequent refreshes mean higher bills.

Documentation

View the full documentation.

CloudWatch data source

MySQL

Tue, 18 Aug 2026 05:43:38 -0700

MySQL: Query your databases directly

What it’s for: Application data including sales metrics, user counts, inventory, and any MySQL-stored data.

Compatible databases

MySQL, MariaDB, Percona Server, Amazon Aurora MySQL, Azure Database for MySQL, Google Cloud SQL for MySQL

Good to know

Query language: SQL
Built into Grafana, with nothing extra to install
A general-purpose database, not a time-series store like Prometheus or InfluxDB

Best for

Application dashboards
Data already in MySQL or compatible databases
Combining technical and application metrics

Tip: MySQL is a general-purpose database rather than a time-series store like Prometheus or InfluxDB. As a general practice, point Grafana at a read replica for complex queries or large datasets so they don’t impact your production database.

Set it up

Open this path in your Grafana Cloud stack for a fully interactive experience, or read through it to understand the process first.

Learning path

Connect to a MySQL data source

Welcome to the Grafana learning path that provides the best practice for configuring a MySQL data source plugin.

11 min

Beginner

Docs & blog posts

Open in Grafana Cloud

Complete this learning path directly in your Grafana Cloud stack with an interactive learning experience.

Open in Grafana Cloud

View on website

PostgreSQL

Tue, 18 Aug 2026 05:43:38 -0700

PostgreSQL: Query your databases directly

What it’s for: Application data including sales metrics, user counts, inventory, JSON data, and any PostgreSQL-stored data.

Also works with

TimescaleDB (time-series extension for PostgreSQL)

Good to know

Query language: SQL, including advanced features like common table expressions (CTEs) and window functions
Native JSON and JSONB support for querying semi-structured data
Works with TimescaleDB, the time-series extension for PostgreSQL

Best for

Application dashboards
Data already in PostgreSQL
Semi-structured data (JSON)
Time-series data with TimescaleDB
Combining technical and application metrics

Tip: As a general practice, point dashboard queries at a read replica to protect a busy production database.

Documentation

View the full documentation.

PostgreSQL data source

Infinity

Tue, 18 Aug 2026 05:43:38 -0700

Infinity: Query any API or file

ℹ Install separately: Infinity is a plugin maintained by Grafana Labs that you install separately. It’s not included with Grafana by default.

Keep it current: On Grafana Cloud, plugins update automatically. On self-managed Grafana, check Plugins periodically so you don’t miss new features or fixes.

What it’s for: External data including REST APIs, JSON, CSV files, XML, and GraphQL endpoints.

Good to know

Queries REST APIs and files, parsing JSON, CSV, XML, and GraphQL responses
Supports multiple authentication methods for connecting to secured APIs
Installed as a plugin, with no separate collector to deploy
Because it queries external APIs, results depend on those APIs’ availability and rate limits, and you often shape the response yourself.

Best for

Third-party APIs (GitHub, Jira, etc.)
CSV/JSON data files
Custom internal APIs
Data not in traditional databases

Example Uses: GitHub stats, weather data, IoT platforms, internal service APIs

Set it up

Open this path in your Grafana Cloud stack for a fully interactive experience, or read through it to understand the process first. This path builds a dashboard from JSON data.

Learning path

Visualize JSON data

Welcome to the Grafana learning path that provides the best practice for creating a dashboard that visualizes JSON data using the Infinity data source.

15 min

Beginner

Docs & blog posts

Open in Grafana Cloud

Complete this learning path directly in your Grafana Cloud stack with an interactive learning experience.

Open in Grafana Cloud

View on website

Working with CSV files instead? Follow the Infinity CSV path.

Infinity: Query CSV data

Google Sheets

Tue, 18 Aug 2026 05:43:38 -0700

Google Sheets: Spreadsheet data in dashboards

What it’s for: Human-maintained data including on-call schedules, manual logs, budgets, and targets.

Good to know

Reference cells or ranges; there’s no query language to learn
Requires Google authentication
Best for small, human-maintained data rather than high-volume telemetry; Google’s sheet size and API rate limits apply

Best for

On-call schedules
Manual tracking and logs
Application targets and KPIs
Data non-engineers maintain

Tip: Reference cells using A1 notation, like Sheet1!A1:E100.

Documentation

View the full documentation.

Google Sheets data source

Troubleshoot a failed connection

Tue, 18 Aug 2026 05:43:38 -0700

When Save & test fails, check these first

A failed connection test usually comes down to a handful of causes:

Check	Why it’s common
URL, port, and protocol (`http` or `https`)	The wrong port or endpoint is a leading cause of connection failures.
Credentials and token scope	Auth errors often look like connection errors. Confirm the account or token has the permissions the data source needs.
Network reachability	Grafana must be able to reach the endpoint over the network. A firewall, VPC, or private network can block access.

Tip: Read the exact error message. A 401 or 403 points to credentials or permissions, while a timeout or “connection refused” points to networking.

Learn more

The data source troubleshooting guide covers connection, authentication, and query errors in detail.

Troubleshoot data sources