Application Real-Time Monitoring Service:Access the Tempo-compatible API of ARMS via Grafana

Prerequisites

Before you begin, make sure you have:

• Grafana 9.x, 10.x, or 11.x installed. Grafana 10.3.0 and later support the Tempo V2 API; earlier versions support only V1. This guide uses Grafana 11.5.2 as an example

• An application monitored by ARMS or Managed Service for OpenTelemetry

• A RAM user with an AccessKey pair that has permissions to query SLS Logstores. For details, see Grant permissions to a RAM user

Add ARMS as a Tempo data source

1. Log on to the Grafana console.

2. In the left-side navigation pane, choose Connections > Add new connection.

3. On the Add new connection page, find and click Tempo.

   

4. Click Add new data source.

   

5. On the Settings tab, configure the Connection and HTTP headers sections as described in the following table, and then click Save & test. Leave all other parameters at their default values.

   Section	Parameter	Description
   Connection	URL	The SLS endpoint in the format https://<sls-endpoint>/trace/tempo/. Replace <sls-endpoint> with the SLS endpoint for the region where your trace data is stored. Example: https://cn-hangzhou.log.aliyuncs.com/trace/tempo/. For the full list of endpoints, see Endpoints.
   HTTP headers	username	The AccessKey ID of a RAM user that has SLS Logstore query permissions. To grant permissions, see Configure the permission assistant feature. To get an AccessKey pair, see AccessKey pair.
   HTTP headers	password	The AccessKey secret that corresponds to the AccessKey ID specified for username.
   HTTP headers	userId	The ID of your Alibaba Cloud account.
   HTTP headers	regionId	The ID of the region where your trace data is stored, such as cn-hangzhou.

   

   

6. Verify the connection. If Successfully connected to Tempo data source appears, the configuration is correct. If an error occurs, see Troubleshoot common issues.

   

Search for traces

After the data source is connected, you can search and inspect traces on the Grafana Explore page.

1. Open the Explore page and select the Tempo data source you created.

2. Filter traces by application name, span name, status, time range, latency, or span attributes.

3. Click a trace ID to open the detail view, where you can diagnose errors and high-latency spans.

Build a trace dashboard

Create dashboard panels to display filtered trace data. The following example builds an error trace list for a specific application.

1. Create or open a dashboard, and then add a new panel.

2. Select the Tempo data source.

3. Expand the search bar and set the following filters:

   • Service Name: the application name.

   • Status: error.

4. Click Save.

Alternatively, enter a TraceQL query directly:

{resource.service.name="mall-gateway" && status=error}

The resulting chart shows all error traces for the specified application: