# Monitoring

This article describes how to use the Monitoring plugin (opens new window).

# Function introduction

The monitoring plug-in is used to monitor the usage of CPU, memory, disk IO, network IO and other hardware resources of the server where the host application is located. The usage of JVM resources, such as heap memory usage, non-heap memory usage, cache usage, throughput (QPS, TPS, average response time). The monitoring plug-in relies on Prometheus for indicator collection. Prometheus will periodically pull the indicator data collected by the monitoring plugin.

# Parameter configuration

# Plugin configuration

The monitoring plugin needs to be configured with monitoring enable switch (monitor.config.enableStartService), IP address/domain name of the host application's environment (monitor.config.address), the port of the host service (monitor.config.port) and the reporting method (monitor.config.reportType). The configuration file of the plug-in can be found in the ${path}/sermant-agent-x.x.x/agent/pluginPackage/monitor/config/config.yaml. The configuration is as follows:

monitor.config:                     # Monitoring plug-in configuration.
    enableStartService: false       # Monitoring plug-in start switch. When it is true, the indicator is collected.
    address: 127.0.0.1              # IP address/domain name of the host application's environment. Used when creating a server instance. Prometheus obtains the indicator information collected by the plug-in by calling the created server instance.
    port: 12345                     # Provides the port information of Http service externally. Used when creating a server instance. Prometheus obtains the indicator information collected by the plug-in by calling the created server instance.
    reportType: PROMETHEUS          # Monitoring indicator reporting method. Currently only PROMETHEUS is supported.
    userName:                       # Authorization information -- authorization username. After the authorization information is configured, prometheus also needs to configure the authorization information to obtain the indicator normally to prevent malicious requests to obtain the indicator information.
    password:                       # Authorization information -- authorized user password -- AES encryption ciphertext. After configuring the authorization information, prometheus also needs to configure the authorization information to obtain the indicator to prevent malicious requests to obtain the indicator information.
    key:                            # KEY used for password encryption of authorized users.
Parameter key Description Default value Required
monitor.config.enableStartService Monitoring plugin start switch false Yes
monitor.config.address IP address/domain name of the host application's environment 127.0.0.1 Yes
monitor.config.port Port information for external Http service 12345 Yes
monitor.config.reportType Monitoring indicator reporting method currently only supports PROMETHEUS PROMETHEUS Yes
monitor.config.userName authorization information -- authorization user name null No
monitor.config.password authorization information -- authorized user password -- AES encryption null No
monitor.config.key KEY used for password encryption of authorized users Empty No

# Detailed governance rules

The indicator data that the monitoring plug-in can collect at present is shown in the following table (after docking with prometheus, users can query the specific indicator information in prometheus through the indicator name, see Verification):

# Resource indicators

Indicator name Description Indicator type
cpu_user proportion of user time CPU
cpu_sys System time share CPU
cpu_wait Percentage of waiting time CPU
cpu_idle idle time ratio CPU
cpu_cores CPU physical cores CPU
memory_total Total memory size memory
memory_swap Size of swap space for cache memory memory
memory_cached Total physical memory of cache memory memory
memory_buffer Make a buffer size for the file memory
memory_used Used memory size memory
disk_readBytesPerSec Number of bytes read by the disk per second during the collection cycle Disk IO
disk_writeBytesPerSec Number of bytes written on the disk per second during the collection cycle Disk IO
disk_ioSpentPercentage Percentage of time spent on disk IO during the acquisition cycle Disk IO
network_readBytesPerSec Number of bytes read by the network per second during the collection cycle Network
network_writeBytesPerSec Number of bytes read by the network per second during the collection cycle Network
network_readPackagePerSec Number of network bytes per second in the collection cycle Network
network_writePackagePerSec Percentage of network IO time spent in the collection cycle Network
heap_memory_init heap memory initialization value JVM
heap_memory_max Maximum heap memory JVM
heap_memory_used Heap memory used JVM
heap_memory_committed The heap memory has been committed JVM
non_heap_memory_init Non-heap memory initialization value JVM
non_heap_memory_max Maximum non-heap memory JVM
non_heap_memory_used Non-heap memory used value JVM
non_heap_memory_committed Non-heap memory committed JVM
code_cache_init Code cache initialization value JVM
code_cache_max Maximum code cache size JVM
code_cache_used The code cache has been used JVM
code_cache_committed The code cache has been committed JVM
meta_sapce_init Metaspace initialization value JVM
meta_sapce_max Maximum value of metaspace JVM
meta_sapce_used Metaspace used value JVM
meta_sapce_committed Metaspace committed value JVM
compressed_class_space_init Compressed class space initialization value JVM
compressed_class_space_max Maximum compressed class space JVM
compressed_class_space_used Compressed class space used value JVM
compressed_class_space_committed Compressed class space committed value JVM
eden_init eden memory initialization value JVM
eden_max maximum memory of eden area JVM
eden_used used value of eden area memory JVM
eden_committed eden area memory committed value JVM
survivor_init memory initialization value of the survivor area JVM
survivor_max maximum memory of the survivor area JVM
survivor_used Memory used value of the survivor area JVM
survivor_committed memory committed value of the survivor area JVM
old_gen_init Memory initialization value in the old era JVM
old_gen_max Maximum memory in the old age JVM
old_gen_used Memory used in the old age JVM
old_gen_committed Memory committed value in the old age JVM
thread_live Active thread JVM
thread_peak thread peak JVM
thread_daemon daemon thread JVM
new_gen_count Young generation GC times JVM
new_gen_spend Young generation GC time JVM
old_gen_count GC times of older generation JVM
old_gen_spend GC time of the elderly generation JVM
cpu_used CPU usage of JVM JVM
start_time JVM started time, milliseconds JVM

# Data Request Metrics

Indicator name Description Indicator type
qps Requests per second Throughput
tps Number of requests processed per second Throughput
avg_response_time Average response time Throughput

# Circuit breaker indicator

Indicator Name Description Indicator Type
fused_request Number of fused requests (number of requests triggering fused matching rules) Fused
failure_fuse_request Number of failed requests (number of requests that failed and ignored exceptions) Breach
failure_rate_fuse_request Request failure rate circuit breaker
avg_response_time average response time circuit breaker
qps Requests per second Blowdown
tps number of transactions processed per second circuit breaker
slow_call_number number of slow calls circuit breaker
permitted_fuse_request number of allowed calls circuit breaker
buffered_calls_number number of cache calls circuit breaker
slow_call_failure_number number of slow call failures circuit breaker
successful_calls_number number of successful calls circuit breaker
not_permitted_calls_number number of calls not allowed circuit breaker
slow_call_success_number number of successful slow calls circuit breaker

# Supported versions and restrictions

Framework support:

  • SpringBoot 1.5.10. Release and above

  • Dubbo 2.6.x-2.7.x

Restrictions:

  • Dependent on Prometheus

  • The collection of server indicators depends on the Linux environment

# Operation and result verification

The following will demonstrate how to use the monitoring plug-in.

# Preparations

Note: Dynamic Configuration Center will be used by default in this scenario. Since it is not the core component of this scenario, it will not be described in this article.

# Step 1: Compile and package the demo application

Execute the following command in the root directory of the demo application to package the demo application:

mvn clean package

After the packaging is successful, a target folder will be generated in the demo root directory, enter the target folder to get the demo-0.0.1-SNAPSHOT.jar package.

# Step 2: Modify the configuration

  • Modify the monitoring plug-in configuration, which can be found in the ${path}/sermant-agent-x.x.x/agent/pluginPackage/monitor/config/config.yaml.
monitor.config:                   # Monitoring plugin configuration
    enableStartService: true      # Monitoring plugin start switch. Modify to true.
    address: 127.0.0.1            # IP address/domain name of the host application's environment. Modify to host specific IP address.
    port: 12345                   # Provides the port information of Http service externally. Modify to an available port.
    reportType: PROMETHEUS        # Monitoring indicator reporting method. Currently only PROMETHEUS is supported.

Add job information under the original job information.

scrape_ configs:
  - job_Name: "prometheus"            # Job name. This is the original task of prometheus.
    static_configs:
      - targets: ["localhost:9090"]   # Monitoring host address
  - job_Name: "Sermant"               # Job name. Add the job of collecting and monitoring plug-in indicators
    metrics_path: /                   # The request path for collecting indicators. Default is/
    basic_auth:                       # Collect the authorization information of indicators, which is consistent with the monitoring plug-in configuration. The monitoring plug-in can be deleted when it is not configured.
      username:                       # Collect the authorization information of the indicator - username
      password:                       # Collect the authorization information of the indicator - password
    static_configs:                   
      - targets: ["127.0.0.1:12345"]  # The host address of the collection index. The IP and port information here is consistent with the monitoring plug-in configuration

# Step 3: Start the application

  • Start the demo application with the following command
# Run under Linux
java -javaagent:${path}/sermant-agent-x.x.x/agent/sermant-agent.jar=appName=default -jar demo-0.0.1-SNAPSHOT.jar

# Run under Windows
java -javaagent:${path}\sermant-agent-x.x.x\agent\sermant-agent.jar=appName=default -jar demo-0.0.1-SNAPSHOT.jar

illustrate: Where path needs to be replaced with the actual installation path of Sermant. x.x.x represents a Sermant version number.

  • Start Prometheus

# Verification

Open Prometheus (default is http://127.0.0.1:9090), query indicators. For example, query heap_memory_used. If the information can be queried, the plug-in will become effective.

The query effect is as follows:

Last Updated: 11/18/2024, 1:42:10 AM

← Service visibility →