Unified visibility across AWS, Azure, and GCP
| Cloud | Metrics | Logs |
|---|---|---|
| AWS | CloudWatch Scrape, CloudWatch Streams | CloudWatch Logs, Firehose |
| Azure | Azure Monitor via Alloy | Event Hub, Azure Functions |
| GCP | Cloud Monitoring via Alloy | Pub/Sub |
Choose your strategy
| Strategy | How it works | Best for |
|---|---|---|
| Store in Grafana Cloud ✓ recommended | Collect metrics/logs → Grafana Cloud storage | Long retention, unified alerting, and access to advanced tools like Knowledge Graph and Grafana Assistant |
| Query directly | Query cloud APIs on demand | Evaluating Grafana with existing data, or strict data residency requirements |
| Hybrid | Store critical, query the rest | When data residency prevents storing some data in Grafana Cloud |
Questions answered
| With Cloud Provider Observability, you can answer… |
|---|
| How do my AWS EC2 instances compare to Azure VMs in one view? |
| Which cloud region has the highest latency? |
| What were my Lambda function errors last month? |
| How is my multi-cloud infrastructure performing overall? |
Problems solved
| Problem | Solution |
|---|---|
| Metrics siloed in each cloud | Unified multi-cloud visibility |
| Limited cloud provider retention | 13 months for metrics in Grafana Cloud |
| Different query languages | PromQL/LogQL for everything |
Script
If you’re running workloads in AWS, Azure, GCP, or across multiple clouds, Cloud Provider Observability is how you bring all that visibility together.
Each cloud provider has its own monitoring system: CloudWatch for AWS, Azure Monitor, Google Cloud Monitoring. They’re fine on their own, but trying to compare EC2 performance against Azure VMs? Or troubleshooting an issue that spans multiple clouds?
That’s where things get difficult.
Cloud Provider Observability solves this by giving you options. You can collect metrics and logs from each cloud and store them in Grafana Cloud — giving you long-term retention, unified alerting, and access to advanced capabilities like Knowledge Graph and Grafana Assistant.
Or you can query each cloud’s native monitoring directly. Quick setup, no data movement, but you miss out on those deeper capabilities.
For most teams, storing in Grafana Cloud is the recommended path. It keeps things simple: you always know where your data is, and you unlock the full platform.
The result? You can answer questions like “Which cloud region has the highest latency?” or “What were my Lambda errors last month?” all from one place, using a consistent query language.