All Products
Search
Document Center

Application Real-Time Monitoring Service:Active profiling

Last Updated:Mar 12, 2024

The active profiling feature of Application Real-Time Monitoring Service (ARMS) Application Monitoring allows you to manually create and start active profiling tasks that listen to calls based on your business requirements.

Important

The active profiling feature is no longer maintained. ARMS provides the hotspot code analysis feature with lower overheads and more accurate method stacks as an alternative. We recommend that you use the hotspot code analysis feature together with the continuous profiling feature for diagnostics. For more information, see Use the hotspot code analysis feature to diagnose slow traces.

Prerequisites

The application is monitored by Application Monitoring. For more information, see Overview.

Note

You must upgrade the ARMS agent to 2.7.1.1 or later. For more information, see Upgrade the ARMS agent.

Create an active profiling task

  1. Log on to the ARMS console. In the left-side navigation pane, choose Application Monitoring > Applications.

  2. On the Applications page, select a region in the top navigation bar and click the name of the application that you want to manage.

    Note

    If the Java图标 icon is displayed in the Language column, the application is connected to Application Monitoring. If a hyphen (-) is displayed, the application is connected to Managed Service for OpenTelemetry.

  3. In the left-side navigation pane, choose Application Diagnosis > Active analysis.

  4. In the New tasks section, set all parameters and click Finish.

    Parameter

    Description

    Interface name

    The operation that you want to perform.

    Thread profiling duration

    The validity period of the task. Valid values:

    • 5min

    • 10min

    • 15min

    Thread profiling trigger threshold (ms)

    The duration that has elapsed before the task starts listening. For example, a slow call consumes 1.4s in total and Thread profiling trigger threshold (ms) is set to 100. In this case, the call is listened to only from 100ms to 1400ms. Valid values: 50 to 30000. Unit: milliseconds.

    Thread profiling frequency (ms)

    The intervals at which API snapshots are generated during thread profiling. The smaller the value, the higher the profiling accuracy, and the higher the performance overhead. We recommend that you set this parameter to 100. Valid values: 5 to 1000. Unit: milliseconds.

    Maximum number of samples

    The maximum number of slow calls that the task can listen to. Valid values: 1 to 100.

View the details of an active profiling task

  1. On the Active analysis page, click View historical tasks.

  2. In the upper-right corner of the page, specify a time range. In the Task list section, click the name of the active profiling task that you want to view.

    After you specify a time range, the Task list section displays all the active profiling tasks created within the time range.

    主动剖析-历史任务

    Click Details in the lower-right corner of the profiling task to view the details of the task.主动剖析-任务详情

    On the Log tab, you can view the logs generated for the task.主动剖析-日志

    In the right-side Thread analysis information section, you can view the information about the API calls that the task listens to, including trace IDs, the points in time when logs were generated, the names of the involved APIs, response codes, IP addresses, and time consumed.

    • You can click the ID of a trace to view the details of the trace. For more information, see Trace details.

    • You can click Thread analysis in the Details column of a trace to view the thread profiling information about the trace.线程剖析