Huawei APM User Manual

Download

APM

User Guide

Issue 01

Date 2021-01-05

HUAWEI TECHNOLOGIES CO., LTD.

No part of this document may be reproduced or transmitted in any form or by any means without prior written consent of Huawei Technologies Co., Ltd.

Trademarks and Permissions

and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd. All other trademarks and trade names mentioned in this document are the property of their respective holders.

Notice

The purchased products, services and features are stipulated by the contract made between Huawei and the customer. All or part of the products, services and features described in this document may not be within the purchase scope or the usage scope. Unless otherwise specied in the contract, all statements, information, and recommendations in this document are provided "AS IS" without warranties, guarantees or representations of any kind, either express or implied.

The information in this document is subject to change without notice. Every eort has been made in the preparation of this document to ensure accuracy of the contents, but all statements, information, and recommendations in this document do not constitute a warranty of any kind, express or implied.

APM User Guide Contents

1 Before You Start....................................................................................................................... 1

2 Permissions Management..................................................................................................... 2

2.1 Creating a User and Granting Permissions.....................................................................................................................2

2.2 Creating a Custom Policy..................................................................................................................................................... 3

3 Application Overview............................................................................................................. 5

3.1 Dashboard..................................................................................................................................................................................5

3.2 Inventory.................................................................................................................................................................................... 6

4 Topology.................................................................................................................................... 7

5 Tracing......................................................................................................................................11

5.1 Call Chain.................................................................................................................................................................................11

5.2 Method Tracing......................................................................................................................................................................14

6 Transactions............................................................................................................................ 16

7 SQL Analysis........................................................................................................................... 19

8 JVM Monitoring..................................................................................................................... 22

9 ICAgent Installation and

9.1 Agent Management (HUAWEI CLOUD Host).............................................................................................................28

9.1.1 Installing the ICAgent (Linux)....................................................................................................................................... 28

9.1.2 Upgrading the ICAgent (Linux).................................................................................................................................... 32

9.1.3 Uninstalling the ICAgent (Linux)................................................................................................................................. 33

9.2 Agent Management (Non-HUAWEI CLOUD Host)...................................................................................................35

9.2.1 Installing the ICAgent...................................................................................................................................................... 35

9.2.2 Upgrading the ICAgent....................................................................................................................................................38

9.2.3 Uninstalling the ICAgent.................................................................................................................................................39

9.3 Collection Conguration.....................................................................................................................................................40

9.4 Conguration Center........................................................................................................................................................... 41

Conguration.......................................................................... 28

10 Alarm Center........................................................................................................................ 43

10.1 Viewing Alarms................................................................................................................................................................... 43

10.2 Viewing Events.................................................................................................................................................................... 46

10.3 Setting Alarm

Notication.............................................................................................................................................. 47

APM User Guide 1 Before You Start

1 Before You Start

This document describes how to use Application Performance Management (APM).

Topology The call and dependency relationships between applications are

displayed, and abnormal instances can be automatically discovered.

Call Chain Information such as the call status, duration, and API is displayed,

helping you further locate fault causes.

TransactionsKey metrics of transactions are displayed and Application

Performance Index (Apdex) values intuitively reect users' satisfaction with applications.

● When a transaction is abnormal, an alarm is reported.

● For transactions with poor user experience, faults can be located based on topology and tracing.

Method Tracing

SQL Analysis

JVM Monitoring

Developers are able to locate method-level performance problems online.

Database performance problems caused by abnormal SQL statements are analyzed. The topology displays the key metrics of databases and SQL statements.

The memory and thread metrics of the JVM running environment are monitored in real time, enabling you to quickly detect problems such as memory leakage and thread exceptions.

● The topology displays the JVM metrics of instances.

● When a JVM metric is abnormal, an alarm is reported.

APM User Guide 2 Permissions Management

2 Permissions Management

2.1 Creating a User and Granting Permissions

This section describes the ne-grained permissions management provided by Identity and Access Management (IAM) for your Application Performance Management (APM). With IAM, you can:

Prerequisites

● Create IAM users for employees based on the organizational structure of your enterprise. Each IAM user has their own security credentials, providing access to APM resources.

● Grant only the permissions required for users to perform a task.

● Entrust a cloud account or service to perform professional and on your APM resources.

If your account does not need individual IAM users, then you may skip over this chapter.

This section describes the procedure for granting permissions (see Figure 2-1).

Before assigning permissions to user groups, you should learn about the APM permissions listed in Permissions Management. For the system permissions of other services, see System Permissions.

ecient O&M

APM User Guide 2 Permissions Management

Process

Figure 2-1 Granting APM permissions

1. Creating a User Group and Assigning Permissions

Create a user group on the IAM console, and assign the APM ReadOnlyAccess policy to the group.

2. Creating an IAM User

Create a user on the IAM console and add the user to the group created in 1.

3. Logging In Using an IAM User and Verifying Permissions

2.2 Creating a Custom Policy

Custom policies can be created as a supplement to the system policies of Application Performance Management (APM). For the actions supported for custom policies, see Permissions Policies and Supported Actions.

You can create custom policies in either of the following two ways:

● Visual editor: Select cloud services, actions, resources, and request conditions without the need to know policy syntax.

● JSON: Edit JSON policies from scratch or based on an existing policy.

For details, see Creating a Custom Policy. The following section contains examples of common APM custom policies.

Example Custom Policies

● Example 1: Allowing a user to install the ICAgent

{ "Version": "1.1",

APM User Guide 2 Permissions Management

"Statement": [ { "Eect": "Allow", "Action": [ "apm:icmgr:create" ] } ] }

● Example 2: Denying collection component uninstallation

A deny policy must be used in conjunction with other policies to take

eect. If the permissions assigned to a user contain both Allow and Deny actions, the Deny actions take precedence over the Allow actions.

To grant a user the APM FullAccess system policy but forbid the user to uninstall collection components, create a custom policy that denies the uninstallation of collection components and grant both the APM FullAccess and deny policies to the user. Because the Deny action takes precedence, the user can perform all operations except uninstalling collection components. The following is an example deny policy:

{ "Version": "1.1", "Statement": [ {

"Eect": "Deny", "Action": [ "apm:icmgr:delete" ] } ] }

● Example 3:

Dening permissions for multiple services in a policy

A custom policy can contain actions of multiple services that are all of the project-level type. The following is an example policy containing actions of multiple services:

{ "Version": "1.1", "Statement": [ {

"Action": [ "aom:*:list", "aom:*:get", "apm:*:list", "apm:*:get" ] }, {

"Action": [ "cce:cluster:get", "cce:cluster:list", "cce:node:get", "cce:node:list" ] } ] }

"Eect": "Allow",

NO TE

APM User Guide 3 Application Overview

3 Application Overview

3.1 Dashboard

An application is a logical group of the same or similar services categorized based on service requirements. You can put services that one application for performance management. For example, you can put accounts, products, and payment services into the Mall application.

You can quickly obtain the health status of applications through the dashboard. On the Dashboard page, you can perform the following operations:

Figure 3-1 Dashboard page

fulll the same function into

The Enterprise Project option is displayed only when you have enabled the enterprise project function. After this function is enabled, both historical and new probe applications are added to the default enterprise project by default. To change the enterprise project to which an application belongs, click Enterprise Project on the application card to go to the enterprise project page and migrate the Application Performance Management (APM) application. Enterprise Project Management Service (EPS) provides a manage cloud resources and personnel by enterprise project. The default project is default. For details about how to enable, create, and manage enterprise projects, see Enterprise

Management User Guide.

You can delete a service card in the following scenarios:

unied method to

APM User Guide 3 Application Overview

● The service connected to APM has been deleted.

● The ICAgent has been uninstalled and service data does not need to be collected.

If the service connected to APM is still running, the service card will be displayed again three minutes after it is deleted.

3.2 Inventory

The Inventory page displays metrics such as the service type, resource ID, response time, calls, and errors, facilitating fault locating.

Figure 3-2 Inventory page

APM User Guide 4 Topology

4 Topology

A topology graphically displays call and dependency relationships between applications. In a topology, each circle represents a service, each segment in a circle represents an instance, and each arrow represents a call relationship. Application Performance Management (APM) supports calls between applications. The topology can display service call relationships across applications. When a circle represents an application, right-click the circle and choose View Application to go to the topology page.

Topology Page

Dierent colors on the circle represent dierent health statuses of instances. Colors are determined by Application Performance Index (Apdex) values. If an Apdex value is closer to 1, the corresponding application is healthier.

1. Table 4-1 provides topology description.

APM User Guide 4 Topology

Table 4-1 Topology description

Color Instance Call

Green 0.75 ≤ Apdex ≤ 1

The instance responds quickly when it is called.

Yellow 0.3 ≤ Apdex < 0.75

The instance responds slowly when it is called.

Red 0 ≤ Apdex < 0.3

The instance responds very slowly when it is called.

Gray The instance is not called. N/A

Black The instance is deleted. N/A

2. On the right of the topology page, set a time range to view the following topology details of an application:

– Transaction Apdex

– Top 5 services ranked by errors and latency

– Top 5 transactions ranked by errors and latency

– Top 5 SQL statements ranked by response time, calls, and errors

3. In the topology, click a circle (a service) to view metric data, including Service Level Agreement (SLA) metrics, basic service metrics, and transaction details.

4. In the topology, click a segment (an instance) in a circle to view metric data, including basic instance metrics, JVM metrics, node metrics, and transaction details.

0.75 ≤ Apdex ≤ 1

Quick response.

0.3 ≤ Apdex < 0.75

Slow response.

0 ≤ Apdex < 0.3

Very slow response.

Locating Faults Using the Topology

The following describes how to locate an instance with a slow response:

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Topology.

Step 3 In the upper right corner of the topology page, set a time range during which a

problem occurs.

Step 4 Check the instance with a long execution time (that is, the instance highlighted in

red) in the topology.

APM User Guide 4 Topology

Step 5 (Optional) For the service containing multiple instances, right-click the service and

choose Expand from the shortcut menu to view call relationships between instances to preliminarily identify the abnormal instance.

Step 6 Choose Find Call-Chain from the shortcut menu. On the page that is displayed,

further locate the fault based on call duration and other parameters.

----End

Conguring Transaction Apdex Threshold

The response time of dierent transactions is dierent. APM enables you to congure dierent Apdex thresholds for dierent transactions. For example, if a

login takes more than 50 ms, the response is slow. If a query transaction takes more than 10 ms, the response is slow. In this case, you need to set Apdex thresholds for the login and query transactions.

Step 1 In the topology page, move the mouse cursor over a circle, right-click it, and click

Edit Threshold.

Step 2 Modify the transaction Apdex threshold and click Apply.

dierent

APM User Guide 4 Topology

----End

APM User Guide 5 Tracing

5 Tracing

5.1 Call Chain

With the tracing function, Application Performance Management (APM) traces and records service calls, comprehensively monitors key metrics such as call status and latency, and visually restores the execution traces and statuses of service requests in distributed systems, so that you can quickly locate performance bottlenecks and faults.

Locating Performance Bottlenecks

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Tracing > Call Chain.

Step 3 In the upper right of the Call Chain page, select the desired time range,

application, and service from three drop-down lists, and click Search.

Step 4 (Optional) On the Call Chain page, click Advanced in the upper right corner, set

lter criteria, and click Search.

Step 5 Identify a service with long call duration and then locate the performance

bottleneck.

Step 6 Click View Call Relationship in the Operation column of the target service.

Step 7 (Optional) View additional information to further locate the cause.

APM User Guide 5 Tracing

On the call relationship page that is displayed, click View Details in the

Operation column to view call details.

----End

Locating Faults

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Tracing > Call Chain.

Step 3 In the upper right of the Call Chain page, select the desired time range,

application, and service from three drop-down lists, and click Search.

Step 4 (Optional) On the Call Chain page, click Advanced in the upper right corner, set

lter criteria, and click Search.

Step 5 Check the service status in the Status column and

nd out the faulty service.

APM User Guide 5 Tracing

Step 6 Click View Call Relationship in the Operation column, check whether the return

value is normal, and locate the fault.

Step 7 (Optional) View additional information to further locate the cause.

On the call relationship page that is displayed, click View Details in the

Operation column to view call details.

----End

NO TE

APM User Guide 5 Tracing

5.2 Method Tracing

Method tracing is used to dynamically trace a method of a class. When the method of this class is called, Application Performance Management (APM) collects the call data of the method based on using probes, and displays the call data on the Call Chain page. Method tracing is used to help application developers locate method-level performance problems online.

APM traces the APIs of most third-party open-source components, but does not

specic methods in your applications. To monitor important methods in

trace applications or methods of some third-party open-source components that are not supported by APM, you need to customize method tracing. After the conguration is complete, you can view the call data of the method on the Call Chain page.

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Tracing > Method Tracing.

Step 3 Customize a method tracing rule and start method tracing.

congured method tracing rules

On the Method Tracing page, click Add Method Tracing Rule, set parameters, and click OK.

● If Method Parameter is not set, all information about the methods using the same name is collected by default.

● If Value is not set, the values of methods are not

● If Collect Method Stack Info is enabled, the call stack information about methods is collected.

● If Collect All Matched Call Info is enabled, all tracing method information is collected. If this function is disabled, tracing method information is collected based on the sampling ratio set during Collection

Conguration.

ltered during collection.

Step 4 Preliminarily locate service performance problems based on the displayed call

duration and status.

APM User Guide 5 Tracing

Step 5 Click View Call Relationship in the Operation column to view the method-level

call relationships.

----End

APM User Guide 6 Transactions

6 Transactions

A transaction is usually an HTTP request. The process is as follows: user request > web server > database > web server > user request. In real life, a transaction is a one-time task. A user completes a task by using an application. In the example of an e-commerce application, querying a commodity is a transaction, and making a payment is also a transaction.

To complete a transaction, you may call multiple services. Any slow or error call may lead to slow responses. During routine O&M, you can analyze the transactions with slow responses to locate and solve application problems, thereby improving user experience.

Transaction Insights Page

Figure 6-1 Transaction insights page

1. Set a time range to view the following transaction details of an application:

– Calls and errors

– Total latency

2. Click Create Group, select a transaction to move it to the new group, and then name the group.

3. Click View Topology to view the topology of the transaction.

4. Click More in the Operation column and select View Call Relationship from the drop-down list to view the tracing data of the transaction.

APM User Guide 6 Transactions

Analyzing Faults Based on Transactions

The following describes how to analyze a transaction with an extremely slow response:

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Transactions.

Step 3 On the Transactions page, select a transaction with an extremely slow response

from the transaction list.

Step 4 Click View Topology in the Operation column to view the topology and instance

details of the transaction.

Step 5 Right-click an instance with an extremely slow response and choose Find Call-

Chain from the shortcut menu. On the page that is displayed, further locate the

fault based on call duration and other parameters.

----End

Customizing Transactions

To precisely customize transactions and classify requests into dierent transactions. When the collector receives requests, custom transactions will be calculated

Step 1 On the Transactions page, click Custom Transaction Rule. A transaction consists

of the request method and regular expression. It is in the format of {Request Method}_/{pattern}. Example: When the request methods are GET and POST and the regular expression is /{name}, the transaction is GET,POST_/{name}.

dene transactions and collect tracing data, use the URI template to

rst.

NO TE

APM User Guide 6 Transactions

Step 2 Select one or more request methods. Request methods include GET, PUT, DELETE,

POST, HEAD, CONNECT, OPTIONS, PATCH, TRACE, and Select all. Select all

indicates all request methods.

Step 3 In the Regular Expression text box, enter a transaction rule and click OK. In this

way, the custom transaction rule is added successfully.

The regular expression uses the URI template matching mode of the Spring MVC framework. Example: @RequestMapping(path="/owners/{ownerId}/pets/{petId}", method=RequestMethod.GET), where ownerId and petId are variables.

To add multiple custom transaction rules, click Add Rule.

● A transaction rule must be 1 to 50 characters long. It must start with a slash (/) but cannot end with a slash. Only letters, digits, and special characters (?*|={}&) are allowed.

● Both the question mark (?) and asterisk (*) can be used for fuzzy search. One question mark represents one character, one asterisk represents 0 to N characters between two slashes in a URI, and double asterisks represent enter /rst/*, /rst/test can be returned but /rst/test/test cannot. When you enter / rst/**, both /rst/test and /rst/test/test can be returned.

innite characters. Example: When you

----End

APM User Guide 7 SQL Analysis

7 SQL Analysis

Application Performance Management (APM) displays key metrics, such as SQL statement calls, response time, and errors for analyzing database performance problems caused by slow or error SQL statements. SQL analysis supports MySQL, Oracle, and PostgreSQL relational databases only.

SQL Page

Figure 7-1 SQL page

Analyzing Abnormal SQL Statements

When an SQL statement of a database is abnormal, performance problems such as service timeout may occur. During routine O&M, you can monitor key metrics, such as error duration and latency of databases, locate the SQL statements that take a long time to execute, operate at low then analyze and optimize them.

The SQL analysis function determines whether to collect SQL data. Before performing the following operations, ensure that this function is enabled.

eciency, or fail to be called, and

APM User Guide 7 SQL Analysis

Otherwise, no SQL data can be queried. This function is enabled by default. If it is disabled, choose Agent > Conguration in the navigation pane and then enable it.

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose SQL Analysis.

Step 3 On the SQL Analysis page, select the time range during which a problem

occurred.

Step 4 On the Overview tab page, locate the faulty database in the application based on

key metrics. If a database requires long response time and has many call errors, performance problems may occur.

Step 5 Analyze the problem cause.

Click the SQL Analysis tab, and locate the abnormal SQL statement in the SQL statement list.

Step 6 Further analyze the cause.

1. Click the abnormal SQL statement to go to the Call Chain page and check the impact of this statement on the entire service.

APM User Guide 7 SQL Analysis

2. Click View Call Relationship in the Operation column to nd out the method of the abnormal SQL statement. Analyze the cause of the abnormal SQL statement in this method. For example, check whether the index is used, data volume is overlarge, syntax is correct, or deadlock occurs. Then, optimize the SQL statement accordingly.

----End

APM User Guide 8 JVM Monitoring

8 JVM Monitoring

JVM monitoring displays the memory and thread metrics of the JVM running environment based on Java applications. You can monitor metric trends in real time for performance analysis.

On the Memory and Thread tab pages, you can respectively view the memory and thread graphs to quickly locate problems such as memory leakage and thread exceptions.

Memory Graphs

As shown in Figure 8-1, in a selected time range, the trends of the maximum, committed, and used memory in memory, heap memory, and non-heap memory spaces) of an instance are displayed. In addition, the garbage collection (GC) duration and times are also displayed.

Figure 8-1 Memory graphs

JVM memory

dierent JVM memory spaces (such as the total

JVM memory consists of heap and non-heap memory.

● Heap memory: A heap is the data area where the JVM is running. It allocates memory for all instances and arrays. Heap memory of objects is reclaimed by an automatic memory management system called garbage collector. Heap space consists of eden space, survivor space, and tenured space.

APM User Guide 8 JVM Monitoring

● Non-heap memory: Memory (excluding heap memory) managed by JVM. Non-heap space consists of code cache and permanent space (or meta space).

Java heap is the main area managed by the garbage collector. It is also called garbage collection heap. GC mode includes full GC and minor GC.

Table 8-1 Memory spaces

Space

Description

Name

Eden

Initially allocates memory from the thread pool to most objects.

space

Survivor

Stores the eden space's objects that are not reclaimed during GC.

space

Tenured space

Code

Maintains objects that have been stored in the survivor space for a period of time.

Compiles and stores local code.

cache

Permanent

Stores static data of VMs, for example, classes and method objects.

space

Meta space

Direct

Stores local class metadata. In versions later than Java 8, permanent space is replaced by meta space.

Resource usage of the direct buer is monitored.

Buer

Full GC Indicates the GC performed in the entire heap space (covering

young-, old-, and permanent-generation spaces) when the memory space is still insucient after memory reclamation.

Minor GC Indicates the GC performed in the young-generation space

(including eden and survivor spaces) when the allocated memory is insucient.

JVM collects garbage based on generations. JVM heap space is divided into oldand young-generation spaces. More than 90% objects that exist only for a short period of time are stored in the young-generation space, whereas objects that have long life cycles are stored in the old-generation space. Young-generation space is further divided into eden space and two survivor spaces. New objects are initially allocated to the eden space. The survivor spaces are used as the

buer

between eden space and tenured space. Objects that are survived after several rounds of GC in the survivor spaces are then transferred to the old-generation space, as shown in Figure 8-2.

NO TE

APM User Guide 8 JVM Monitoring

Figure 8-2 Memory spaces

Thread Graphs

There are two survivor spaces, which are represented by from and to pointers. The to pointer points to the empty survivor space.

As shown in Figure 8-3, in a selected time range, the trends of new, runnable, blocked, and waiting threads are displayed.

Figure 8-3 Thread graphs

APM User Guide 8 JVM Monitoring

Table 8-2 Threads

Thread Name

Total threads Both active and standby threads are included. Sticky threads

Deadlock threads

New threads Number of threads that are newly created.

Runnable threads

Blocked threads

Waiting threads

Timed waiting threads

Description

and dedicated threads become standby threads after being executed.

When two or more threads encounter resource conicts or the communication between them is abnormal, the system enters the deadlock state.

Number of threads that can run.

Number of threads that are blocked.

Number of threads that are in the waiting state.

Number of threads that are waiting for another thread to perform an action for a specied waiting time.

Terminated threads

Max connections

Current connections

Max threads Maximum number of threads that can be executed on Tomcat.

Current threads

Busy threads Number of threads executed on Tomcat for processing tasks.

Adding a Threshold Rule

You can add threshold rules for all JVM memory and thread metrics. When the rules are met, alarms are reported, altering you to risks.

Step 1 On the JVM Monitoring page, select an application in the upper left corner, and

then select an instance.

Number of threads that are terminated.

Maximum number of connections that are supported by Tomcat.

Number of connections that are being occupied on Tomcat.

Number of threads that are being executed on Tomcat.

Step 2 In the trend graph of a memory or thread metric on the right, set a threshold rule.

Specically, click Add Threshold Rule on the top of the trend graph.

NO TE

APM User Guide 8 JVM Monitoring

Step 3 Set rule parameters and click Submit, as shown in the following gure. If you

want to receive alarm notications, select Yes when setting Send Notication and then select a topic.

Description of the Add For Service parameter:

● If this parameter is set to Yes, the threshold rule is applied to the entire service.

● If this parameter is set to No, the threshold rule is applied to a single instance.

APM User Guide 8 JVM Monitoring

----End

APM User Guide 9 ICAgent Installation and Conguration

9 ICAgent Installation and Conguration

9.1 Agent Management (HUAWEI CLOUD Host)

9.1.1 Installing the ICAgent (Linux)

Prerequisites

Before installing the ICAgent, ensure that the time and time zone of the local browser are consistent with those of the desired server. If multiple servers are deployed, ensure that the local browser and multiple servers use the same time zone and time. Otherwise, the application topology and tracing data on the console may be incorrect.

Installation Methods

There are two methods to install the ICAgent. The two methods are not applicable to container nodes created using ServiceStage, or Cloud Container Engine (CCE). To monitor container nodes through Application Performance Management (APM), see APM Getting Started. Table 9-1 lists the ICAgent installation methods.

Table 9-1 Installation methods

Method

Initial installation

Application Scenario

This installation method is used when the following conditions are met:

1. An Elastic IP Address (EIP) has been bound to the server. For details, see Assigning an EIP and Binding It to an ECS.

2. The ICAgent has never been installed on the server.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Method Application Scenario

Inherited installation

Initial Installation

After you apply for a server on the cloud and install the ICAgent in the Linux environment, perform the following operations:

Step 1 Obtain an Access Key ID/Secret Access Key (AK/SK) by using either of the

following methods:

● Obtain a temporary AK/SK by creating an agency. For details, see How Do I

This installation method is used when the following conditions are met:

You have multiple servers on which the ICAgent is to be installed. One server is bound to an EIP, but others are not bound to an EIP. The ICAgent has been installed on the server bound to an EIP. You can use this method to install the ICAgent on the servers that are not bound to an EIP.

Obtain the AK/SK by Creating an Agency?.

For each ECS server where the ICAgent is to be installed, you need to bind it to an agency on the ECS console. The agency relationship takes eect 5 minutes later.

● Obtain a permanent AK/SK by adding access keys. For details, see How Do I

Obtain the AK/SK and Project ID?.

Step 2 Log in to the APM console. In the navigation pane, choose Agent > Management.

Step 3 Click Install ICAgent. On the page that is displayed, set Host to HUAWEI CLOUD

host and OS to Linux.

Step 4 Generate the ICAgent installation command and copy it.

● As shown in Figure 9-1, if you have obtained the permanent AK/SK, set Installation Mode to Obtain AK/SK and enter the AK/SK in the text box to generate the ICAgent installation command. Then, click Copy Command.

Figure 9-1 Entering the AK/SK

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Ensure that the AK/SK are correct. Otherwise, the ICAgent cannot be installed.

● If you have obtained the temporary AK/SK, set Installation Mode to Create

Agency and click Copy Command to copy the ICAgent installation command.

Step 5 Use a remote login tool to log in to the server where the ICAgent is to be installed

as the root user and run the preceding command to install the ICAgent.

● If the message "ICAgent install success" is displayed, the ICAgent is successfully installed in the /opt/oss/servicemgr/ directory. After the ICAgent is successfully installed, choose Agent > Management in the navigation pane to view the ICAgent status.

● If the installation fails, uninstall the ICAgent according to Uninstalling the ICAgent

(Linux) and then install it again. If the problem persists, contact technical support.

----End

Inherited Installation

If the ICAgent has been installed on a server and the installation package ICProbeAgent.tar.gz exists in the /opt/ICAgent/ directory of the server, use this method to install the ICAgent on a remote server with a few clicks.

Step 1 Run the following command (

x.x.x.x

indicates the server IP address) on the server

where the ICAgent has been installed:

bash /opt/oss/servicemgr/ICAgent/bin/remoteInstall/remote_install.sh -ip x.x.x.x

Step 2 Enter the password of the root user of the server where the ICAgent is to be

installed as prompted.

● If both the expect tool and ICAgent have been installed on a server, the ICAgent is successfully installed on the remote server after the preceding command is run. If the ICAgent has been installed on a server, but the expect tool has not, enter the information as prompted.

● Ensure that the root user can run the SSH and SCP commands on the ECS server where the ICAgent has been installed to communicate with the remote ECS server where the ICAgent is to be installed.

● If the installation fails, uninstall the ICAgent according to Uninstalling the ICAgent

(Linux) and then install it again. If the problem persists, contact technical support.

----End

Inherited Batch Installation

NO TICE

NO TE

APM User Guide 9 ICAgent Installation and Conguration

1. Ensure that you can run the SSH and SCP commands on the ECS server where the ICAgent has been installed to communicate with the remote ECS servers where the ICAgent is to be installed.

2. If you have installed the ICAgent in a server through an agency, you also need to set an agency for other servers where the ICAgent is to be installed. For details, see Creating an Agency.

3. Batch installation scripts depend on Python versions. You are advised to implement batch installation on hosts running Python 2.x. Python 3.x does not support batch installation.

Prerequisites

The IP addresses and passwords of all ECS servers where the ICAgent is to be installed have been collected, sorted in the iplist.cfg le, and uploaded to the /opt/ICAgent/ directory on the ECS server where the ICAgent has been installed. As shown in the following example, each IP address and password in the iplist.cfg

le must be separated by a space.

192.168.0.109 password

192.168.0.39 password

● Because the iplist.cfg le contains sensitive information, you are advised to clear it in time.

● If the passwords of all servers are the same, you only need to list IP addresses in the iplist.cfg le and enter the password once during execution. If the password of an IP address is dierent from those of other IP addresses, you need to list both passwords and IP addresses in the iplist.cfg

● The batch installation function depends on Python 2.7.*. If the system displays a message indicating that Python cannot be found during the installation, install Python

2.7.* and try again.

(Enter the actual password.)

le.

Procedure

Step 1 Run the following command on the server where the ICAgent has been installed:

bash /opt/oss/servicemgr/ICAgent/bin/remoteInstall/remote_install.sh -

batchModeCong

/opt/ICAgent/iplist.cfg

Enter the default password of the root user of the servers where the ICAgent is to be installed as prompted. If the passwords of all IP addresses have been congured in the iplist.cfg le, press Enter to skip this step. Otherwise, enter the default password.

batch install begin Please input default passwd: send cmd to 192.168.0.109 send cmd to 192.168.0.39 2 tasks running, please wait... 2 tasks running, please wait... 2 tasks running, please wait... End of install agent: 192.168.0.39 End of install agent: 192.168.0.109 All hosts install icagent

nish.

APM User Guide 9 ICAgent Installation and Conguration

Wait until the message "All hosts install icagent nish." is displayed, which indicates that the ICAgent has been successfully installed on all the hosts listed in the conguration le.

Step 2 After the ICAgent is successfully installed, choose Agent > Management in the

navigation pane to view the ICAgent status.

----End

ICAgent Statuses

The following table lists the ICAgent statuses.

Table 9-2 ICAgent statuses

Status Description

Running The ICAgent is running properly.

Uninstalled The ICAgent is not installed. For details about how to install the

ICAgent, see Installing the ICAgent (Linux).

Installing The ICAgent is being installed. This operation takes about 1

minute to complete.

Installation failed

Upgrading The ICAgent is being upgraded. This operation takes about 1

Upgrade failed

Oine The AK/SK or ECS agency congurations are incorrect. Ensure

Faulty The ICAgent is faulty. Contact technical support.

Failed to install the ICAgent. Uninstall the ICAgent according to

Uninstalling the ICAgent Through Logging In to a Server

and then install it again.

minute to complete.

Failed to upgrade the ICAgent. Uninstall the ICAgent according to Uninstalling the ICAgent Through Logging In to a Server and then install it again.

that such congurations are correct.

9.1.2 Upgrading the ICAgent (Linux)

To ensure better collection experience, Application Performance Management (APM) will continuously upgrade ICAgent versions. When the Linux system displays a message indicating that a new ICAgent version is available, perform the following operations:

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Agent > Management.

Step 3 Select Cluster: XXX or Other:

right of the page.

user-dened nodes from the drop-down list on the

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Step 4 Upgrade the ICAgent.

● If you select Cluster: xxx in Step 3, directly click Upgrade ICAgent. In this way, the ICAgent on all hosts in the cluster can be upgraded at a time.

● If you select Other: then click Upgrade ICAgent.

Step 5 In the displayed Upgrade ICAgent dialog box, click Yes. Wait for about 1 minute

to complete the ICAgent upgrade. When the ICAgent status changes from

Upgrading to Running, the ICAgent is successfully upgraded.

----End

user-dened nodes in Step 3, select a desired host and

9.1.3 Uninstalling the ICAgent (Linux)

If the ICAgent on a server is uninstalled, server O&M will be topology and tracing functions unavailable. Exercise caution when performing this operation.

You can uninstall the ICAgent using either of the following methods:

● Uninstalling the ICAgent Through the APM Console: The ICAgent has been successfully installed, and needs to be uninstalled.

● Uninstalling the ICAgent Through Logging In to a Server: The ICAgent fails to be installed, and needs to be uninstalled.

● Remotely Uninstalling the ICAgent: The ICAgent has been successfully installed, and needs to be remotely uninstalled.

● Uninstalling the ICAgent in Batches: The ICAgent has been successfully installed, and needs to be uninstalled in batches.

Uninstalling the ICAgent Through the APM Console

Step 1 Log in to the Application Performance Management (APM) console. In the

navigation pane, choose Agent > Management.

Step 2 Select Other:

page.

Step 3 Select one or more servers whose ICAgent is to be uninstalled, and click Uninstall

ICAgent. In the Uninstall ICAgent dialog box, click Yes.

user-dened nodes from the drop-down list on the right of the

aected, making

Wait for about 1 minute to complete the uninstallation. When the ICAgent status changes from Uninstalling to Uninstall, the ICAgent is successfully uninstalled.

To reinstall the ICAgent, wait for 5 minutes after it is uninstalled. Otherwise, the ICAgent may be automatically uninstalled.

----End

Uninstalling the ICAgent Through Logging In to a Server

Step 1 Log in to the server from which the ICAgent is to be uninstalled as the root user.

Step 2 Run the following command to uninstall the ICAgent:

NO TE

NO TICE

APM User Guide 9 ICAgent Installation and Conguration

bash /opt/oss/servicemgr/ICAgent/bin/manual/uninstall.sh;

Step 3 Wait until the message "ICAgent uninstall success" is displayed.

----End

Remotely Uninstalling the ICAgent

In addition to the preceding methods, you can use a method similar to Inherited

Installation to remotely uninstall the ICAgent.

Step 1 Run the following command (

where the ICAgent has been installed:

bash /opt/oss/servicemgr/ICAgent/bin/remoteUninstall/remote_uninstall.sh ip x.x.x.x

Step 2 Enter the password of the root user of the server where the ICAgent is to be

uninstalled as prompted.

● If both the expect tool and ICAgent have been installed on a server, the ICAgent is successfully uninstalled from the remote server after the preceding command is run. If the ICAgent has been installed on a server, but the expect tool has not, enter the information as prompted.

● Ensure that the root user can run the SSH and SCP commands on the Elastic Cloud Server (ECS) server where the ICAgent has been installed to communicate with the remote ECS server where the ICAgent is to be uninstalled.

● If the message "ICAgent uninstall success" is displayed, the ICAgent is successfully uninstalled. After the ICAgent is successfully uninstalled, choose Agent > Management in the navigation pane to view the ICAgent status.

----End

Uninstalling the ICAgent in Batches

x.x.x.x

indicates the server IP address) on the server

If the ICAgent has been installed on a server and the installation package ICProbeAgent.zip exists in the /opt/ICAgent/ directory of the server, use this method to uninstall the ICAgent from multiple remote ECS servers with a few clicks.

The ECS servers must belong to the same Virtual Private Cloud (VPC) and network segment.

Prerequisites

The IP addresses and passwords of all ECS servers where the ICAgent is to be uninstalled have been collected, sorted in the iplist.cfg

le, and uploaded to the /opt/ICAgent/ directory on the ECS server where the ICAgent has been installed. As shown in the following example, each IP address and password in the iplist.cfg

192.168.0.109 password

le must be separated by a space.

(Enter the actual password.)

NO TE

APM User Guide 9 ICAgent Installation and Conguration

192.168.0.39 password

● Because the iplist.cfg le contains sensitive information, you are advised to clear it in time.

dierent from those of other IP addresses, you need to list both passwords

(Enter the actual password.)

Procedure

Step 1 Run the following command on the server where the ICAgent has been installed:

bash /opt/oss/servicemgr/ICAgent/bin/remoteUninstall/remote_uninstall.sh batchModeCong /opt/ICAgent/iplist.cfg

Enter the default password of the root user of the servers where the ICAgent is to be uninstalled as prompted. If the passwords of all IP addresses have been congured in the iplist.cfg le, press Enter to skip this step. Otherwise, enter the default password.

batch uninstall begin Please input default passwd: send cmd to 192.168.0.109 send cmd to 192.168.0.39 2 tasks running, please wait... End of uninstall agent: 192.168.0.109 End of uninstall agent: 192.168.0.39 All hosts uninstall icagent

nish.

Wait until the message "All hosts uninstall icagent nish." is displayed, which indicates that the ICAgent has been successfully uninstalled from all the hosts listed in the

conguration le.

Step 2 After the ICAgent is successfully uninstalled, choose Agent > Management in the

navigation pane to view the ICAgent status.

----End

9.2 Agent Management (Non-HUAWEI CLOUD Host)

9.2.1 Installing the ICAgent

Prerequisites

● You have purchased an Elastic Cloud Server (ECS) as a jump server.

● The operating system (OS) of the server meets the requirements in

Supported OSs and supports the AMD64 processor architecture.

● The server has been bound to an Elastic IP Address (EIP). For details, see

Assigning an EIP and Binding It to an ECS.

● The time and time zone of the local browser are consistent with those of the ECS server.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Procedure

Before installing the ICAgent on a non-HUAWEI CLOUD host, purchase an ECS server from HUAWEI CLOUD as a jump server and perform the following operations:

You are advised to use CentOS 6.5 64bit or later images. The minimum specication is

1vCPUs | 1GB and the recommended one is 2vCPUs | 4GB.

Step 1 Log in to the ECS server and modify its security group rule.

1. On the ECS details page, click the Security Groups tab.

2. On the security group page, click a security group name and click Modify Security Group Rule.

3. On the security group details page, click Inbound Rules and then Add Rule. On the page that is displayed, add a security group rule based on Table 9-3.

Table 9-3 Security group rule

Direction Protocol Port Description

Inbound TCP 8149, 8102, 8923,

30200, 30201, and 80

List of ports on the jump server to which the ICAgent sends data

Enable ports 8149, 8102, 8923, 30200, 30201, and 80 in the inbound direction of the security group to ensure normal data communication between the non-HUAWEI CLOUD host and the jump server.

Step 2 In the navigation pane, choose Agent > Management.

Step 3 Select Other: user-dened nodes, click Install ICAgent, and set Host to Non-

HUAWEI CLOUD host.

Step 4 Enable forwarding ports on the jump server.

1. As shown in Figure 9-2, enter the private IP address of the jump server to generate a forwarding command.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Figure 9-2 Entering the private IP address of the jump server

The private IP address of the jump server refers to the internal IP address of the Virtual Private Cloud (VPC) where the jump server locates.

2. Click Copy Command to copy the forwarding command.

3. Log in to the jump server as the root user, and run the SSH tunnel forwarding command:

ssh -f -N -L {ECS IP address}:8149:{ELB IP address}:8149 -L {ECS IP address}:8102:{ELB IP address}: 8102 -L {ECS IP address}:8923:{ELB IP address}:8923 -L {ECS IP address}:30200:{ELB IP address}:30200 L {ECS IP address}:30201:{ELB IP address}:30201 -L {ECS IP address}:80:icagent-{Region}.obs. {Region}.myhuaweicloud.com:80 {ECS IP address}

Enter the password of the root user as prompted.

4. Run the netstat -lnp | grep ssh command to check whether corresponding ports are being listened to. If the results in Figure 9-3 are returned, TCP ports are enabled.

Figure 9-3

Verication results of TCP ports

– Enter http://ECS IP address in the address bar of the browser. If the access is

successful, the security group rule has taken eect.

– If the jump server is powered o and then restarted, run the preceding command

again.

Step 5 Obtain an Access Key ID/Secret Access Key (AK/SK). For details, see How Do I

Obtain the AK/SK and Project ID?.

Step 6 Generate the ICAgent installation command and copy it.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

1. As shown in Figure 9-4, enter the AK, SK, DC, and Connection IP to generate the ICAgent installation command.

Figure 9-4 Entering the AK/SK

– Ensure that the AK/SK are correct. Otherwise, the ICAgent cannot be installed.

– DC: Custom a DC name for querying hosts more easily.

– Connection IP: For EIP connection, use the EIP of the jump server. For VPC peer

connection, use the internal IP address of the VPC where the jump server locates.

2. Click Copy Command.

Step 7 Use a remote login tool to log in to the server where the ICAgent is to be installed

as the root user and run the preceding command to install the ICAgent.

If the message "ICAgent install success" is displayed, the ICAgent is successfully installed in the /opt/oss/servicemgr/ directory. After the ICAgent is successfully installed, choose Agent > Management in the navigation pane to view the ICAgent status.

----End

9.2.2 Upgrading the ICAgent

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Agent > Management.

Step 3 Select Cluster: XXX or Other:

user-dened nodes from the drop-down list on the

right of the page.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Step 4 Upgrade the ICAgent.

● If you select Cluster: xxx in Step 3, directly click Upgrade ICAgent. In this way, the ICAgent on all hosts in the cluster can be upgraded at a time.

● If you select Other: user-dened nodes in Step 3, select a desired host and then click Upgrade ICAgent.

Step 5 In the displayed Upgrade ICAgent dialog box, click Yes. Wait for about 1 minute

to complete the ICAgent upgrade. When the ICAgent status changes from

Upgrading to Running, the ICAgent is successfully upgraded.

----End

9.2.3 Uninstalling the ICAgent

If the ICAgent on a server is uninstalled, server O&M will be aected, making topology and tracing functions unavailable. Exercise caution when performing this operation.

You can uninstall the ICAgent using either of the following methods:

● Uninstalling the ICAgent Through the APM Console: The ICAgent has been successfully installed, and needs to be uninstalled.

● Uninstalling the ICAgent Through Logging In to a Server: The ICAgent fails to be installed, and needs to be uninstalled.

Uninstalling the ICAgent Through the APM Console

Step 1 Log in to the Application Performance Management (APM) console. In the

navigation pane, choose Agent > Management.

Step 2 Select Other:

page.

Step 3 Select one or more servers whose ICAgent is to be uninstalled, and click Uninstall

ICAgent. In the Uninstall ICAgent dialog box, click Yes.

Wait for about 1 minute to complete the uninstallation. When the ICAgent status changes from Uninstalling to Uninstall, the ICAgent is successfully uninstalled.

To reinstall the ICAgent, wait for 5 minutes after it is uninstalled. Otherwise, the ICAgent may be automatically uninstalled.

----End

user-dened nodes from the drop-down list on the right of the

Uninstalling the ICAgent Through Logging In to a Server

Step 1 Log in to the server from which the ICAgent is to be uninstalled as the root user.

Step 2 Run the following command to uninstall the ICAgent:

bash /opt/oss/servicemgr/ICAgent/bin/manual/uninstall.sh;

Step 3 Wait until the message "ICAgent uninstall success" is displayed.

----End

NO TE

APM User Guide 9 ICAgent Installation and Conguration

9.3 Collection Conguration

To reduce memory, database, and disk space usage, you can implement collection conguration as required. The collection conguration takes eect for selected applications.

Procedure

Step 1 Log in to the Application Performance Management (APM). In the navigation

pane, choose Agent >

Step 2 Select an application from the Application drop-down list.

Conguration.

Step 3 Click

This function is enabled by default. When you do not need to collect tracing and topology data of a specic application, disable this function to reduce resource usage.

to enable data collection.

Step 4 Click to enable the function of collecting normal call chain data.

To reduce the resources consumed by probes, APM collects one more data record every minute when a transaction is abnormal or the latency is greater than

Application Performance Index (Apdex) Threshold. If this function is enabled,

normal call chain data is sampled and collected. If this function is disabled, normal call chain data is not collected.

Step 5 Click

to enable memory monitoring.

To prevent probes from aecting service performance in peak hours, enable memory monitoring. When the instance memory usage is excessively high, probes

enter the hibernation state. You can also click to set the duration and memory usage.

● Memory usage = Used memory of the Java process/Maximum available memory

● Maximum available memory: Use the smaller value between the available memory quota of the container and the maximum heap memory of the JVM. The maximum heap memory of the JVM is the value of -Xmx. The default value is 25% of the maximum available memory of the JVM.

● The memory usage during collection suspension must be greater than or equal to that during collection restoration.

Step 6 Click

to enable the function of adding trace IDs to logs.

A trace ID uniquely identies a tracing. When this function is enabled, the system adds trace IDs to logs. You can accurately search for logs based on trace IDs, such

e1c08cab, e1c08cad, and e1c08cae, as shown in the

as following gure.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

Step 7 Click to enable SQL analysis.

When this function is disabled, no SQL data is aected, but you cannot implement SQL analysis.

Step 8 Set the HTTP response codes to be ignored.

To prevent unnecessary error reporting, and quickly and accurately locate faulty tracing, set the HTTP response codes to be ignored. They will not be recorded in

error record tables. Click click . If there are multiple HTTP response codes, separate them by commas (,).

Step 9 Set the errors and exceptions to be ignored.

To prevent unnecessary error reporting, and quickly and accurately locate faulty tracing, set the errors and exceptions to be ignored. They will not be recorded in

error record tables. Click click . If there are multiple types of Java exceptions, separate them by commas (,). The default value is null.

, enter the HTTP response codes to be ignored, and

, enter the errors and exceptions to be ignored, and

----End

9.4 Conguration Center

Setting Apdex Thresholds

Step 1 Log in to the Application Performance Management (APM) console. In the

navigation pane, choose

Step 2 Select an application from the drop-down list.

Step 3 Set Application Performance Index (Apdex) thresholds. For details, see Basic

Concepts.

● Click threshold, and click to save the threshold.

● Click Apdex threshold, and click to save the threshold.

next to Topology Apdex Threshold (ms), enter a topology Apdex

The default topology Apdex threshold is 100 ms.

next to Transaction Apdex Threshold (ms), enter a transaction

Conguration Center.

NO TE

APM User Guide 9 ICAgent Installation and Conguration

– The default transaction Apdex threshold is 500 ms.

– This setting takes eect for all transactions of the application. If an Apdex

threshold has been separately set for a transaction, the currently set Apdex threshold takes an Apdex threshold for a transaction, do as follows:

1. In the navigation pane, choose Transactions.

2. In the drop-down list in the upper left corner, select the application to which the transaction belongs.

3. In the transaction list, click desired transaction, enter an Apdex threshold, and then click

threshold.

eect for all transactions except this transaction. To separately set

in the Apdex Threshold (ms) column of the

to save the

----End

APM User Guide 10 Alarm Center

10 Alarm Center

10.1 Viewing Alarms

Flowchart

Alarms are reported when Application Performance Management (APM) or an external service, such as ServiceStage, or Cloud Container Engine (CCE), is abnormal or may cause exceptions. Alarms need to be handled. Otherwise, service exceptions may occur.

Procedure

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Alarm Center > Alarm List.

APM User Guide 10 Alarm Center

Step 3 View alarms on the Alarm List page.

1. Set a time range to view alarms. There are two methods to set a time range:

Method 1: Use the predened time label, for example, Last 1 hour, Last 6 hours, or Last 1 day. You can choose one based on service requirements.

Method 2: Customize a time range. The time range can be 30 days at most.

2. Set

Step 4 Perform the operations described in Table 10-1 if needed.

Table 10-1 Operations

Operation Method Description

Viewing alarm statistics

Clearing alarms

Viewing alarm details

lter criteria and click Search to view alarms.

Click Reset to reset lter criteria.

View alarm statistics that meet specic lter criteria within a specic time range through a bar graph.

Click Clear in the Operation column of a target alarm.

Click View Details in the Operation column of the target alarm to view alarm details.

● You can clear an alarm after the problem that causes this alarm is resolved.

● Cleared alarms cannot be queried.

Viewing the latest alarms

----End

Click on the right of the page to view the latest three alarms.

Setting Threshold Rules for Transaction Metrics

APM supports alarm reporting when transaction exceptions occur. You can set threshold rules for transaction metrics. When metric values meet the threshold rules, alarms will be reported to the alarm center. To view alarms, choose Alarm Center > Alarm List in the navigation pane. The following uses the Calls metric of transactions as an example.

Step 1 On the Transactions page, choose More > Modify Threshold in the Operation

column. The Add Threshold page is displayed.

NO TE

APM User Guide 10 Alarm Center

Figure 10-1 Adding a threshold rule for a transaction metric

Step 2 Select the Calls metric. Transaction metrics include Calls, Total latency, Errors,

and Apdex.

Step 3 Set the threshold condition: When there are 15 or more calls in 3 minutes, an

alarm will be reported.

Step 4 If you do not need to receive

notications, select Disabled for Notication and

then click OK.

If you want to receive notications, select Enabled. For details, see Setting Alarm

Notication.

----End

Setting Threshold Rules for JVM Metrics

APM supports alarm reporting when JVM memory and thread metrics are abnormal. You can set threshold rules for JVM metrics. When metric values meet the threshold rules, alarms will be reported to the alarm center. To view alarms, choose Alarm Center > Alarm List in the navigation pane. The following uses the

Maximum memory metric of the total memory as an example.

Step 1 On the JVM Monitoring page, click Add Threshold Rule.

NO TE

APM User Guide 10 Alarm Center

Figure 10-2 Adding a threshold rule

Step 2 On the page that is displayed, select the Maximum memory metric name. Other

options are Committed memory and Used memory.

Step 3 Set the threshold condition: When the total memory is greater than or equal to 15

MB for 3 minutes, an alarm will be reported.

Step 4 Select the Average statistical method. Other options are Maximum and

Minimum.

Step 5 Select the Minor alarm severity. Other options are Critical, Major, and Warning.

Step 6 If you do not need to receive

click Submit.

If you want to receive notications, select Yes. For details, see Setting Alarm Notication.

----End

10.2 Viewing Events

Events generally carry some important information. They are reported when Application Performance Management (APM) or an external service, such as

notications, select No for Send Notication and

APM User Guide 10 Alarm Center

ServiceStage, or Cloud Container Engine (CCE) encounters some changes. Such changes do not necessarily cause service exceptions. Events do not need to be handled.

Procedure

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Alarm Center > Event List.

Step 3 View events on the Event List page.

1. Set a time range to view events. There are two methods to set a time range:

Method 1: Use the hours, or Last 1 day. You can choose one based on service requirements.

Method 2: Customize a time range. The time range can be 30 days at most.

2. Set

Step 4 Perform the operations described in Table 10-2 if needed.

Table 10-2 Operations

Operation

Viewing event statistics

----End

lter criteria and click Search to view events.

Click Reset to reset lter criteria.

predened time label, for example, Last 1 hour, Last 6

Method Description

View event statistics that meet lter criteria within a specic time range through a bar graph.

10.3 Setting Alarm Notication

Application Performance Management (APM) supports alarm notication. That is, a certain type of alarms can be sent to (SMS) message or email. In this way, they can identify and rectify cluster exceptions at the earliest time, avoiding service loss.

You can create a maximum of 10 rules reaches 10, delete unnecessary notication rules and create new ones.

If you do not create any notication rule, you cannot receive alarm notications. In that case, you can only log in to the APM console and choose Alarm Center > Alarm List to view alarms.

APM enables you to create notication rules only for the alarms listed in Table

10-3.

specied users by Short Message Service

notication rules. If the number of notication

NO TE

APM User Guide 10 Alarm Center

Table 10-3 Alarm types

Alarm Type Description

Probe hibernation alarm

Collector installation alarm

Threshold alarm

More types of alarms are being developed.

Creating a Notication Rule

Step 1 Log in to the APM console.

Step 2 In the navigation pane, choose Alarm Center >

Create Notication Rule.

Step 3 Create a topic,

have made such congurations, skip this step.

congure a topic policy, and add subscribers to the topic. If you

Generated when the probe is in hibernation state.

Generated when the ICAgent fails to be installed, upgraded, or uninstalled, or is abnormal.

Generated when a threshold rule is triggered.

Notication Rules, and click

1. Access the Simple Message Notication (SMN) console: When APM is interconnected with SMN, click Create SMN Topic to access the SMN console.

2. Create a topic: In the navigation pane of the SMN console, choose Topic Management > Topics. Then click Create Topic. On the page that is displayed, enter a topic name and click OK. For details, see Creating a Topic.

Congure a topic policy according to Figure 10-3. Otherwise, alarm

3. notications will fail to be sent. Then, add subscribers, that is, SMS message or email receivers of alarm

notications. For details, see Conguring Topic

Policies, and Adding a Subscription. In this way, when an exception occurs in

a cluster, APM can broadcast the alarm information to the subscribers in real time.

APM User Guide 10 Alarm Center

Figure 10-3 Conguring a topic policy

Step 4 Enter a rule name, select an alarm type (for details, see Table 10-3), select the

topic created in Step 3, customize a cluster to be monitored, and click Create, as shown in Figure 10-4.

After the generated, APM automatically sends

Figure 10-4 Creating a

notication rule is created, if an alarm that meets the notication rule is

notications by SMS message or email.

notication rule

----End

APM User Guide 10 Alarm Center

More Operations

After creating a notication rule, you can also perform the operations described in

Table 10-4.

Table 10-4 Related operations

Operation Description

Modifying a

notication

rule

Deleting a

notication

rule

Searching for a notication rule

Click Modify in the Operation column.

● To delete a notication rule, click Delete in the Operation column.

● To delete one or more notication rules, select it or them and click Delete above the notication rule list.

Enter a keyword of the notication rule name in the search box

in the upper right corner and click

Huawei APM User Manual

Specifications and Main Features

Frequently Asked Questions

User Manual

Contents

1 Before You Start

2 Permissions Management

2.1 Creating a User and Granting Permissions

2.2 Creating a Custom Policy

3 Application Overview

3.1 Dashboard

3.2 Inventory

4 Topology

5 Tracing

5.1 Call Chain

5.2 Method Tracing

6 Transactions

7 SQL Analysis

8 JVM Monitoring

9.1 Agent Management (HUAWEI CLOUD Host)

9.1.1 Installing the ICAgent (Linux)

9.1.2 Upgrading the ICAgent (Linux)

9.1.3 Uninstalling the ICAgent (Linux)

9.2 Agent Management (Non-HUAWEI CLOUD Host)

9.2.1 Installing the ICAgent

9.2.2 Upgrading the ICAgent

9.2.3 Uninstalling the ICAgent

10 Alarm Center

10.1 Viewing Alarms

10.2 Viewing Events