Integrations vs. data sources
When to use each
| Use integrations when… | Use data sources when… |
|---|---|
| You want fast time-to-value | You already have monitoring infrastructure |
| You need pre-built alerts and dashboards | Compliance requires data to stay on-premises |
| You want Machine Learning insights and SLOs | You just need visualization of existing data |
The key insight: Data source connections give you visualization. Integrations give you the full Grafana Cloud platform.
Script
This is the single most important thing to understand in this course.
When you use an integration, data gets collected and stored in Grafana Cloud. Because the data lives in Grafana Cloud, you get everything the platform offers: pre-built dashboards designed by experts, managed alerting with Machine Learning-powered anomaly detection, SLOs with error budget tracking, and the ability to correlate across all your signals.
When you use a data source connection, data stays wherever it already lives. Grafana queries it on demand for visualization, and that’s it.
No pre-built dashboards, no managed alerts, no Machine Learning insights, no SLOs. Just visualization.
Both have their place. Data sources are great when you already have monitoring infrastructure you don’t want to migrate.
But if you want the full Grafana Cloud experience (the capabilities that make observability powerful), integrations are the path.