All Products
Search
Document Center

Application Real-Time Monitoring Service:Trace query

Last Updated:Dec 27, 2023

On the Trace Query page, you can query traces by using multiple filter conditions or query the details of a specific trace based on the trace ID. You can also aggregate multiple traces for analysis.

Query traces

  1. Log on to the ARMS console.
  2. In the left-side navigation pane, choose Application Monitoring > Trace Query. In the top navigation bar, select a region.

  3. On the Trace Query page, select a parameter from the Parameter Type drop-down list, specify a custom value in the Parameter Value field, and then click Add to Query Condition.

    You can set the Parameter Type parameter to TraceId for exact match.

    Table 1. Parameters

    Parameter

    Description

    TraceId

    Enter a trace ID.

    Operation Name

    Enter an API name. Fuzzy match is not supported.

    Client Application Name

    Enter the name of an application on the client.

    Server Application Name

    Enter the name of an application on the server.

    Duration Exceeds

    Specify a number of milliseconds to query calls that take longer than the specified number of milliseconds.

    Call Type

    Select a call type.

    Abnormal Call

    Set this parameter to true to search for all traces that contain abnormal calls.

    Thread Profiling Snapshots Only

    Set this parameter to true to search for all traces that contain thread profiling snapshots.

    Client IP Address

    Enter the IP address of an application that initiates the call.

    Server IP Address

    Enter the IP address of an application that is called.

    Business Primary Key

    Enter a business primary key to search for business events.

    Response Code

    Enter a response code.

  4. Click the ID of the trace that you want to view. You are redirected to the Traces tab.

    Traces

    The following section describes the parameters on the Traces tab:

    • Application: the name of the application to which the trace belongs.

    • Log Generated At: the time when the log was generated.

    • Status: the status of the trace. Red indicates that an exception exists in the on-premises trace called by the service. Green indicates that the trace is normal.

    • IP Address: the IP address of the application.

    • Call Type: the type of the call, which corresponds to the call type of the ad hoc query.

    • Interface Name: the name of the API that is called.

    • Details: Click the Details icon icon to view information about the method stack, analysis overview, and metric monitoring of the trace. For more information, see Trace details.

    • Timeline (ms): the time consumed by each service to call the trace and the proportion of time consumed by each service in the time consumed to call the entire trace.

Analyze traces

On the Trace Query page, select all the traces that you want to analyze and click Analyze Selected Traces.

In the Trace Analysis panel, you can view the span name, application name, call type, number of requests, proportion of requests, number of exceptions, proportion of exceptions, average self duration, and average duration of all the selected traces.

Trace aggregation for analysis

  • Move the pointer over a span name to view the ID of the trace that contains the span.

  • Click an application name to go to the Application Overview page of the application.

  • Click Statistical Analysis in the Actions column to view details of the span. The details include the proportions of different call types for each API, total duration, API name, number of requests, recommended samples, number of times that each call method is used. The name, type, total duration, and proportion of time consumed by each call method are also included. Details of a span

Trace details

On the Traces tab, click the Details icon icon in the Details column to view the information about the method stack, analysis overview, and metric monitoring of the trace.

Method stack

Method stack

The Method Stack tab contains the following parameters:

  • Methods: the method used to call the on-premises method stack. After the call method is expanded, the next-layer calls of the method are displayed.

  • Line: the line where the code of the method used to call the on-premises method stack is located.

  • Expanded Info:

    • Parameter: the input parameters of the call.

    • SQL: the SQL statements to access data in the database.

    • Exception: the details of the exception.

  • Timeline (ms): the distribution of time consumed by each method call of the on-premises trace.

Analysis overview

On the Analysis Overview tab, you can view the details of the application, including the proportion of call types, average duration, total duration, API name, and number of requests. The name, type, number of calls, execution time, and percentage of execution time of each method are also displayed. Analysis overview

Metric monitoring

On the Indicator monitoring tab, you can view the time series curves for the instantaneous GC count, instantaneous GC time consumption, number of requests, response time, and number of errors. For more information, see JVM monitoring. Metric monitoring