Grafana supports a wide range of data sources, including Prometheus, MySQL, and even Datadog. There's a good chance you can already visualize metrics from the systems you have set up. In some cases, though, you already have an in-house metrics solution that you'd like to add to your Grafana dashboards. Grafana Data Source Plugins enables integrating such solutions with Grafana.
- Grafana >= 10.4.0
- A VTEX account
- The VTEX IO apps you want to monitor must be running on builder-node 7.x
The VTEX IO Grafana Datasource plugin is currently in Closed Beta and distributed as a signed zip file attached to each GitHub Release. Since it isn't published on the official Grafana plugin catalog yet, it must be installed manually.
For the full step-by-step guide — downloading the plugin, installing it on Linux/macOS/Windows/Docker, and configuring the datasource with your VTEX App Key and Token — see docs/PLUGIN_SETUP_CLOSED_BETA.md.
Once the plugin is listed in the Grafana catalog:
- In Grafana, go to the Plugins section and search for VTEX IO.
- Click Install plugin.
- With the plugin installed, click Add new data source.
- Fill in the fields with your VTEX account data:
- App Key: your VTEX API app key
- Account: your VTEX account name (lowercase letters and numbers only)
- App Token: your VTEX API app token
- Click Save & test.
The VTEX IO Grafana datasource provides predefined metric types for easy observability of your VTEX IO applications:
- Select Query Type: Choose "Metrics" from the query type dropdown
- Select App Name: Choose the VTEX IO app you want to monitor
- Select Metric Type: Choose from the predefined metrics:
- Request Rate per Account: Shows the total number of requests over time, grouped by account
- Request Rate per Status per Account: Shows request count broken down by HTTP status code and account
- Latency Stats per Account and Handler: Shows latency percentiles (p50, p95, p99 in ms) per account and handler in a table
All metrics use the runtime_http_requests_total metric from the VTEX Observability Platform:
| Metric Type | Description | Grouping | Backend Metric |
|---|---|---|---|
| Request Rate per Account | Total requests over time | By account | runtime_http_requests_total |
| Request Rate per Status per Account | Requests by HTTP status | By account + status code | runtime_http_requests_total |
| Latency Stats per Account and Handler | Latency percentiles (p50, p95, p99) per account and handler | By account + handler (table) | runtime_http_requests_duration_milliseconds |
Latency Stats — table visualization: Results for "Latency Stats per Account and Handler" are intended to be viewed as a Table (columns: account, handler, p50, p95, p99 in ms). The plugin signals this to Grafana via the response metadata. When you add a new panel and run a Latency Stats query, Grafana will often suggest or default to the Table visualization. In Explore or when changing an existing panel’s query to Latency Stats, if the view does not switch to Table automatically, choose Table from the visualization picker (panel options).
- Select Query Type: Choose "Logs" from the query type dropdown
- Select App Name: Choose the VTEX IO app whose logs you want to view
- Configure Page Size: Adjust the number of log entries to retrieve (default: 100)
Found a bug or want to request a feature? Open a support ticket at supporticket.vtex.com/support.