Managed Service for OpenTelemetry:Quick start

Overview

The trace details page of the legacy console displays the full trace of an HTTP 404 request. The timeline on the left shows five hierarchical spans: GET /** (404, 11.1ms), ResourceHttpRequestHandler.handleRequest (85μs), ResponseFacade.sendError (19μs), BasicErrorController.errorHtml (4.92ms), and Render error (4.34ms). The top section shows the TraceID, start time, total duration, and a red error indicator. The Span Details panel on the right shows the application name java-agent-usage-0.0.1-SNAPSHOT, the interface name BasicErrorController.errorHtml, the spanId, parentSpanId, and attributes such as ali.trace.component_type=98, ali.trace.flag=x-trace, and thread.name=http-nio-8081-exec-1.

This guide uses the OpenTelemetry Java agent with a Java application as an example:

1. Download a demo application: Download the demo to try Managed Service for OpenTelemetry, or instrument your own application.

2. Instrument your application with the OpenTelemetry Java agent: Download the OpenTelemetry Java agent for non-intrusive automatic instrumentation across hundreds of Java frameworks. Modify JVM startup parameters to report traces.

3. View monitoring data: View telemetry data in the Managed Service for OpenTelemetry console or the .

Step 1: Download the demo application

Use the provided demo application or your own Java application.

Download the demo code: OpenTelemetry Java Agent Demo.

Step 2: Instrument the application

The OpenTelemetry Java agent provides non-intrusive automatic instrumentation for hundreds of Java frameworks listed in Supported Libraries and Versions.

1. Download the OpenTelemetry Java Agent.

   In the Assets section of the latest release, download the opentelemetry-javaagent.jar file.

2. Modify the VM parameters in your Java startup command to report traces.

   -javaagent:/{user.workspace}/opentelemetry-javaagent.jar    // Replace {user.workspace} with the actual path where you downloaded the file.
   -Dotel.resource.attributes=service.name=<appName>     // Replace <appName> with your application name.
   -Dotel.exporter.otlp.headers=Authentication=<token>    // Use the token from the endpoint information obtained in the prerequisites.
   -Dotel.exporter.otlp.endpoint=<endpoint>  // Use the endpoint from the endpoint information obtained in the prerequisites.

   For example, if your startup command is:

   java -jar /{user.workspace}/java-agent-usage-0.0.1-SNAPSHOT.jar

   To connect to Managed Service for OpenTelemetry, update the command to:

   java -Dotel.resource.attributes=service.name=java-agent-usage-0.0.1-SNAPSHOT -Dotel.exporter.otlp.headers=Authentication=ggxw4l****@7323a5caae30263_ggxw4l****@53df7ad2afe**** -Dotel.exporter.otlp.endpoint=http://tracing-analysis-dc-hz.aliyuncs.com:8090 -javaagent:/{user.workspace}/opentelemetry-javaagent.jar -jar /{user.workspace}/java-agent-usage-0.0.1-SNAPSHOT.jar

Step 3: View monitoring data

1. Access the demo application at localhost:8081/user/async. The agent collects and sends telemetry data to Managed Service for OpenTelemetry.

2. On the Managed Service for OpenTelemetry console, go to the Applications page, select your application to view traces.

   On the application details page, click the Trace Analysis tab to view charts for the number of calls, errors, and average latency. The trace list shows details such as TraceID, Interface Name, Application Name, Latency, Status, Start Time, and Host Address. Use Slow Trace Analysis and Error Trace Analysis to identify factors that cause slow calls.