Query data in place
| Data source | What you can query |
|---|---|
| Prometheus | Query your existing Prometheus servers directly |
| CloudWatch | Query AWS metrics without storing in Grafana Cloud |
| Azure Monitor | Direct queries to Azure metrics and logs |
| Elasticsearch | Visualize existing log data |
| InfluxDB | Query time-series data in place |
| SQL databases | Query PostgreSQL, MySQL, BigQuery directly |
Questions answered
| With Data Source Connections, you can answer… |
|---|
| How do metrics from my existing Prometheus compare to CloudWatch? |
| What’s happening across all my monitoring tools in one view? |
| Can I correlate my Elasticsearch logs with Prometheus metrics? |
| What does my data look like without storing it in Grafana Cloud? |
Problems solved
| Problem | How data source connections help |
|---|---|
| Don’t want to store data in Grafana Cloud | Query data where it lives |
| Already invested in other monitoring tools | Keep using them, visualize in Grafana |
| Data residency requirements | Data never leaves your infrastructure |
| Need unified view of disparate systems | Single pane of glass for all data sources |
| Data sources on private network | Private Data Source Connect (PDC) for secure access |
Architecture
Your data never moves. Perfect for when you can’t store data in Grafana Cloud or you want to evaluate Grafana with existing data.
Script
The third approach is different: instead of moving data to Grafana Cloud, you query it where it already lives.
Data source connections let you connect to your existing Prometheus servers, your CloudWatch account, Azure Monitor, Elasticsearch, InfluxDB, SQL databases, you name it. The data never moves. Grafana just queries it on demand.
This is perfect when storing data in Grafana Cloud isn’t an option. Maybe you’ve already invested heavily in another monitoring tool. Maybe you have data residency requirements that prevent data from leaving your infrastructure. Or maybe you just want to see what Grafana can do with your existing data before committing to anything.
The important thing to understand is that this is a visualization approach, not a storage approach. You get the unified view, seeing all your data in one place, but you don’t get Grafana Cloud’s managed storage benefits like long-term retention or recording rules. The data stays yours, in your systems.
One more thing: by default, Grafana Cloud connects over the public internet. If your data sources are on a private network, you’ll need Private Data Source Connect to reach them securely.