Three approaches to data

Here’s something critical to understand before we go further. Data sources aren’t the only way to work with data in Grafana, and they have important trade-offs.

You have three main approaches. Integrations are the “easy button”: install one, and Grafana handles collection, storage, pre-built dashboards, and pre-configured alerts. Everything’s managed.

Data sources are what we’re covering in this course. They connect to data where it already lives. No migration, no new storage. You can query, visualize, and create alerts with data sources. Some even include pre-built dashboards.

But here’s the catch: your data stays external, which means you miss out on some Grafana Cloud features like ML-powered anomaly detection, SLOs, and cross-signal correlations.

The third option is Send Telemetry. You push data directly to Grafana Cloud using agents or SDKs, getting full access to all features.

So when should you use data sources? When you already have existing monitoring infrastructure you don’t want to migrate, when data sovereignty requires data to stay in your systems, or when you’re combining Grafana with existing tools.