Overview
Application Performance Management (APM) is an application performance management platform. Based on real-time multi-language application probe full-scale collection technology, it offers distributed application performance analysis and fault self-detection capability, providing an all-round guarantee for system availability and stability. It assists in quickly locating performance issues in complex business systems and reducing MTTR (Mean Time to Repair). With real-time understanding and tracking of application performance, it helps enhance user experience.
APM supports various mainstream programming languages and open-source frameworks, providing a one-stop solution for application performance monitoring.
-This section introduces the APM API interfaces, which comply with the OpenAPI 3.0 specification.
-You can call the API to perform operations on Application Performance Monitoring (APM), such as creating a business system, retrieving report information, and viewing monitoring metrics.
-For information on all APIs supported by APM, see API Overview.
Glossary
For common terms of the APM API interface, see the table below:
| Term | Description |
|---|---|
| The business system is used to categorize application management. Each business system has a unique Token, which must be specified when accessing the application. You can set storage duration, billing mode, and other parameters at the business system level. It also enables permission management and cost allocation based on the business system. Monitoring data is completely isolated between different business systems. | |
| An application is the most important entity. Multiple processes accessed with the same application name are displayed as multiple instances under the same application. Therefore, an application is a logical combination, which can be equal to a service in microservices architecture, containing multiple peer-to-peer instances. | |
| Service, referring to an independent service unit in a microservices architecture, is usually synonymous with "application". | |
| An instance is an actual deployment unit in a physical or virtual environment, usually corresponding to a process. | |
| Distributed link Tracing is a technology used to monitor and analyze the path of request circulation in Distributed systems. It primarily tracks the entire process of a request from initiation to final processing, recording the calling relationships, execution time, and status between services and components, thereby helping locate bottlenecks, troubleshoot issues, and optimize system reliability. | |
| Span | A span is the minimum unit of work in a distributed tracing system, recording the execution of a single operation (such as making a service call, executing a database query, or sending a message). |
| Span ID | Span ID is the unique identifier of each Span, associated with the parent Span through Parent ID to form a hierarchy. |
| Trace is a directed acyclic graph (DAG) composed of a group of associated Spans, representing the complete call path of a request in the entire system. | |
| Trace ID | Trace uses a globally unique Trace ID to connect all related Spans for cross-service end-to-end performance analysis. |
| The probe (Agent) is a core data collection component responsible for collecting performance data from the application runtime environment and sending it to the monitoring backend. | |
| Event tracking is the process of inserting monitoring logic into user applications to collect performance data. Typically, probes provide automatic event tracking solutions that are non-intrusive to business code. Meanwhile, users can also perform manual event tracking via SDK. | |
| OpenTelemetry is an open source distributed tracing standard that defines APIs and data models, supporting cross-tool compatibility. It also provides instrumentation solutions for mainstream programming languages via probes or SDKs. | |
| Metrics are quantitative measurements of application performance at specific time points or within certain periods, typically stored as time-series data, supporting aggregation calculations (such as average, maximum value, percentile) and trend analysis. |
Usage limits
-By default, Sub-users have no access permission to Application Performance Monitoring (APM). Therefore, we need to assign APM access permission to Sub-users through Cloud Access Management (CAM). For details, see Access Management.
-APM offers a 15-day trial period for new users, with a reporting limit of 100 million Spans and a fixed storage duration of 7 days. If the limit is exceeded or the free trial expires, please refer to the purchase method to enable the official version, otherwise application data reporting will be limited.
Getting Started with APIs
You can use the API Explorer tool to call APIs online.
This document uses creating a business system, obtaining access information, and retrieving observable data as examples. The steps to make an API call via the API Explorer Tool are as follows:
- Go to the API Explorer tool page. For more information about API Explorer tool usage, see Using API Explorer.
- Call the Create APM business system API to create a business system. After creation, the business system ID is returned.
- Call the Get APM Access Point API to obtain access information.
- See the Access Guide for application connection.
- Call the general interface to obtain metric data to get application performance metric data.
Feedback