Managed Service for OpenTelemetry integration guide - Cloud Monitor

Network Analysis and Monitoring integrates with Managed Service for OpenTelemetry to provide end-to-end request tracing. By connecting synthetic monitoring data with backend traces, you can see the nodes, call stacks, and response times along each request path and quickly pinpoint performance issues.

Background

Tracing provides tools to visualize complete call chains, analyze request volumes, map service topologies, and understand application dependencies, helping you quickly diagnose performance bottlenecks in distributed architectures.

Enable this feature to easily connect your synthetic monitoring requests to Managed Service for OpenTelemetry. For more information, see What is Managed Service for OpenTelemetry.
A trace record captures the processing details of a single request, including service calls and their durations.
A trace is the complete record of a request. It represents the path of a transaction or process through a distributed system and is a directed acyclic graph (DAG) of spans. Each span is a named and timed segment of work within the trace. For more information about traces, see Key concepts.

Billing

Network Analysis and Monitoring does not charge for tracing. If you enable Managed Service for OpenTelemetry, you are billed based on your usage of that service.

Procedure

You can enable tracing only when you create HTTP(S) site monitoring tasks for PC and mobile detection points. To enable tracing, specify the data-reporting region and the tracing protocol. Network Analysis and Monitoring supports the OpenTelemetry W3C, Zipkin B3, and Jaeger protocols. Turn on the Enable Tracing switch, select a region such as China (Hangzhou) from the Tracing Data Reporting Address drop-down list, and select a protocol type such as OpenTelemetry W3C from the Tracing Protocol drop-down list. To check whether the service is activated, you can click Check if Managed Service for OpenTelemetry is activated.

The following table lists the regions that support tracing.

Area	Region
Chinese mainland	China (Hangzhou), China (Shanghai), China (Qingdao), China (Beijing), China (Zhangjiakou), China (Hohhot), China (Ulanqab), China (Shenzhen), China (Heyuan), China (Guangzhou), China (Chengdu), and China (Hong Kong)
Asia Pacific	Singapore, Malaysia (Kuala Lumpur), Japan (Tokyo), and Indonesia (Jakarta)
Europe and Americas	UK (London), Germany (Frankfurt), US (Virginia), and US (Silicon Valley)

To connect synthetic monitoring data with your backend systems, the region and protocol you select must match those configured on your backend. The following table lists the tracing protocols that each backend component type supports.

Category	Component	Guide	Supported protocols
Gateway	MSE	Enable tracing for a cloud-native gateway	W3C, B3, SkyWalking
	ACK Ingress	Enable tracing for NGINX Ingress Controller	W3C, B3, Jaeger
	ALB	Configure end-to-end tracing for Application Load Balancer	B3
	ASM	Enable distributed tracing in ASM	W3C, B3
	API Gateway	Configure trace analysis	B3
Backend application	Java/Go/Python (self-developed)	Application integration	W3C, B3, Jaeger, SkyWalking, EagleEye
Backend application	.NET, Node.js, and other languages (open source)	Integration guide	W3C, B3, Jaeger, SkyWalking
Cloud service	Alibaba Cloud Model Studio		W3C
Dependent component	More than 100 plug-ins are supported, covering RPC, message queues, databases, task scheduling, and more.

View traces

On the detection history page for your site monitoring task, select the Normal filter tag to view normal detection records. The table includes columns such as Time, Detection Point, Detection Source, Node Type, Resolved IP, Detection Target, Response Time, and Error Message. To view details about a single detection, click Details in the Actions column.
Select the End-to-end Calls tab to view the trace associated with the detection request. Click a trace to open the details page, where you can view the detailed backend trace. For more information, see Trace analysis. The details page is accessible from the End-to-end Calls tab and provides several analysis tools. At the top, you can find basic information and search by serviceName or traceId. The left pane allows filtering by status, duration, application, and API name. The center displays charts for call volume and duration, while the Span List below shows call details. To the right, the Slow Trace Analysis pane provides recommendations for traces lasting over 500 ms. The details page also shows basic information like traceId, Start Time, and Total Duration, along with a summary of Applications and APIs involved. A waterfall chart visualizes the call chain for each span, detailing the API name (such as GET /user/async), source, role, and duration. Use this chart to locate performance bottlenecks or exceptions in the trace.