你居然还去服务器上捞日志,搭个日志收集系统难道不香么!

Hollis

共 9638字,需浏览 20分钟

 ·

2020-06-20 23:46

ELK日志收集系统进阶使用,本文主要讲解如何打造一个线上环境真实可用的日志收集系统。有了它,你就可以和去服务器上捞日志说再见了!

ELK环境安装

ELK是指Elasticsearch、Kibana、Logstash这三种服务搭建的日志收集系统,具体搭建方式可以参考《SpringBoot应用整合ELK实现日志收集》。这里仅提供最新版本的docker-compose脚本和一些安装要点。

docker-compose脚本

version: '3'
services:
  elasticsearch:
    image: elasticsearch:6.4.0
    container_name: elasticsearch
    environment:
      - "cluster.name=elasticsearch" #设置集群名称为elasticsearch
      - "discovery.type=single-node" #以单一节点模式启动
      - "ES_JAVA_OPTS=-Xms512m -Xmx512m" #设置使用jvm内存大小
      - TZ=Asia/Shanghai
    volumes:
      - /mydata/elasticsearch/plugins:/usr/share/elasticsearch/plugins #插件文件挂载
      - /mydata/elasticsearch/data:/usr/share/elasticsearch/data #数据文件挂载
    ports:
      - 9200:9200
      - 9300:9300
  kibana:
    image: kibana:6.4.0
    container_name: kibana
    links:
      - elasticsearch:es #可以用es这个域名访问elasticsearch服务
    depends_on:
      - elasticsearch #kibana在elasticsearch启动之后再启动
    environment:
      - "elasticsearch.hosts=http://es:9200" #设置访问elasticsearch的地址
      - TZ=Asia/Shanghai
    ports:
      - 5601:5601
  logstash:
    image: logstash:6.4.0
    container_name: logstash
    environment:
      - TZ=Asia/Shanghai
    volumes:
      - /mydata/logstash/logstash.conf:/usr/share/logstash/pipeline/logstash.conf #挂载logstash的配置文件
    depends_on:
      - elasticsearch #kibana在elasticsearch启动之后再启动
    links:
      - elasticsearch:es #可以用es这个域名访问elasticsearch服务
    ports:
      - 4560:4560
      - 4561:4561
      - 4562:4562
      - 4563:4563

安装要点

  • 使用docker-compose命令运行所有服务:

docker-compose up -d
  • 第一次启动可能会发现Elasticsearch无法启动,那是因为/usr/share/elasticsearch/data目录没有访问权限,只需要修改/mydata/elasticsearch/data目录的权限,再重新启动;

chmod 777 /mydata/elasticsearch/data/
  • Logstash需要安装json_lines插件。

logstash-plugin install logstash-codec-json_lines



分场景收集日志

这里为了方便我们查看日志,提出一个分场景收集日志的概念,把日志分为以下四种。

  • 调试日志:最全日志,包含了应用中所有DEBUG级别以上的日志,仅在开发、测试环境中开启收集;

  • 错误日志:只包含应用中所有ERROR级别的日志,所有环境只都开启收集;

  • 业务日志:在我们应用对应包下打印的日志,可用于查看我们自己在应用中打印的业务日志;

  • 记录日志:每个接口的访问记录,可以用来查看接口执行效率,获取接口访问参数。



Logback配置详解

要实现上面的分场景收集日志,主要通过Logback的配置来实现,我们先来了解下Logback的配置吧!

完全配置

在SpringBoot中,如果我们想要自定义Logback的配置,需要自行编写logback-spring.xml文件,下面是我们这次要使用的完全配置。



<configuration>
    
    <include resource="org/springframework/boot/logging/logback/defaults.xml"/>
    
    <include resource="org/springframework/boot/logging/logback/console-appender.xml"/>
    
    <springProperty scope="context" name="APP_NAME" source="spring.application.name" defaultValue="springBoot"/>
    
    <property name="LOG_FILE_PATH" value="${LOG_FILE:-${LOG_PATH:-${LOG_TEMP:-${java.io.tmpdir:-/tmp}}}/logs}"/>
    
    <springProperty name="LOG_STASH_HOST" scope="context" source="logstash.host" defaultValue="localhost"/>

    
    <appender name="FILE_DEBUG"
              class="ch.qos.logback.core.rolling.RollingFileAppender">

        
        <filter class="ch.qos.logback.classic.filter.ThresholdFilter">
            <level>DEBUGlevel>
        filter>
        <encoder>
            
            <pattern>${FILE_LOG_PATTERN}pattern>
            <charset>UTF-8charset>
        encoder>
        <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
            
            <fileNamePattern>${LOG_FILE_PATH}/debug/${APP_NAME}-%d{yyyy-MM-dd}-%i.logfileNamePattern>
            
            <maxFileSize>${LOG_FILE_MAX_SIZE:-10MB}maxFileSize>
            
            <maxHistory>${LOG_FILE_MAX_HISTORY:-30}maxHistory>
        rollingPolicy>
    appender>

    
    <appender name="FILE_ERROR"
              class="ch.qos.logback.core.rolling.RollingFileAppender">

        
        <filter class="ch.qos.logback.classic.filter.LevelFilter">
            <level>ERRORlevel>
            <onMatch>ACCEPTonMatch>
            <onMismatch>DENYonMismatch>
        filter>
        <encoder>
            
            <pattern>${FILE_LOG_PATTERN}pattern>
            <charset>UTF-8charset>
        encoder>
        <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
            
            <fileNamePattern>${LOG_FILE_PATH}/error/${APP_NAME}-%d{yyyy-MM-dd}-%i.logfileNamePattern>
            
            <maxFileSize>${LOG_FILE_MAX_SIZE:-10MB}maxFileSize>
            
            <maxHistory>${LOG_FILE_MAX_HISTORY:-30}maxHistory>
        rollingPolicy>
    appender>

    
    <appender name="LOG_STASH_DEBUG" class="net.logstash.logback.appender.LogstashTcpSocketAppender">
        <filter class="ch.qos.logback.classic.filter.ThresholdFilter">
            <level>DEBUGlevel>
        filter>
        <destination>${LOG_STASH_HOST}:4560destination>
        <encoder charset="UTF-8" class="net.logstash.logback.encoder.LoggingEventCompositeJsonEncoder">
            <providers>
                <timestamp>
                    <timeZone>Asia/ShanghaitimeZone>
                timestamp>
                
                <pattern>
                    <pattern>
                        {
                        "project": "mall-tiny",
                        "level": "%level",
                        "service": "${APP_NAME:-}",
                        "pid": "${PID:-}",
                        "thread": "%thread",
                        "class": "%logger",
                        "message": "%message",
                        "stack_trace": "%exception{20}"
                        }
                    pattern>
                pattern>
            providers>
        encoder>
        
        <connectionStrategy>
            <roundRobin>
                <connectionTTL>5 minutesconnectionTTL>
            roundRobin>
        connectionStrategy>
    appender>

    
    <appender name="LOG_STASH_ERROR" class="net.logstash.logback.appender.LogstashTcpSocketAppender">
        <filter class="ch.qos.logback.classic.filter.LevelFilter">
            <level>ERRORlevel>
            <onMatch>ACCEPTonMatch>
            <onMismatch>DENYonMismatch>
        filter>
        <destination>${LOG_STASH_HOST}:4561destination>
        <encoder charset="UTF-8" class="net.logstash.logback.encoder.LoggingEventCompositeJsonEncoder">
            <providers>
                <timestamp>
                    <timeZone>Asia/ShanghaitimeZone>
                timestamp>
                
                <pattern>
                    <pattern>
                        {
                        "project": "mall-tiny",
                        "level": "%level",
                        "service": "${APP_NAME:-}",
                        "pid": "${PID:-}",
                        "thread": "%thread",
                        "class": "%logger",
                        "message": "%message",
                        "stack_trace": "%exception{20}"
                        }
                    pattern>
                pattern>
            providers>
        encoder>
        
        <connectionStrategy>
            <roundRobin>
                <connectionTTL>5 minutesconnectionTTL>
            roundRobin>
        connectionStrategy>
    appender>

    
    <appender name="LOG_STASH_BUSINESS" class="net.logstash.logback.appender.LogstashTcpSocketAppender">
        <destination>${LOG_STASH_HOST}:4562destination>
        <encoder charset="UTF-8" class="net.logstash.logback.encoder.LoggingEventCompositeJsonEncoder">
            <providers>
                <timestamp>
                    <timeZone>Asia/ShanghaitimeZone>
                timestamp>
                
                <pattern>
                    <pattern>
                        {
                        "project": "mall-tiny",
                        "level": "%level",
                        "service": "${APP_NAME:-}",
                        "pid": "${PID:-}",
                        "thread": "%thread",
                        "class": "%logger",
                        "message": "%message",
                        "stack_trace": "%exception{20}"
                        }
                    pattern>
                pattern>
            providers>
        encoder>
        
        <connectionStrategy>
            <roundRobin>
                <connectionTTL>5 minutesconnectionTTL>
            roundRobin>
        connectionStrategy>
    appender>

    
    <appender name="LOG_STASH_RECORD" class="net.logstash.logback.appender.LogstashTcpSocketAppender">
        <destination>${LOG_STASH_HOST}:4563destination>
        <encoder charset="UTF-8" class="net.logstash.logback.encoder.LoggingEventCompositeJsonEncoder">
            <providers>
                <timestamp>
                    <timeZone>Asia/ShanghaitimeZone>
                timestamp>
                
                <pattern>
                    <pattern>
                        {
                        "project": "mall-tiny",
                        "level": "%level",
                        "service": "${APP_NAME:-}",
                        "class": "%logger",
                        "message": "%message"
                        }
                    pattern>
                pattern>
            providers>
        encoder>
        
        <connectionStrategy>
            <roundRobin>
                <connectionTTL>5 minutesconnectionTTL>
            roundRobin>
        connectionStrategy>
    appender>

    
    <logger name="org.slf4j" level="INFO"/>
    <logger name="springfox" level="INFO"/>
    <logger name="io.swagger" level="INFO"/>
    <logger name="org.springframework" level="INFO"/>
    <logger name="org.hibernate.validator" level="INFO"/>

    <root level="DEBUG">
        <appender-ref ref="CONSOLE"/>
        
        
        <appender-ref ref="LOG_STASH_DEBUG"/>
        <appender-ref ref="LOG_STASH_ERROR"/>
    root>

    <logger name="com.macro.mall.tiny.component" level="DEBUG">
        <appender-ref ref="LOG_STASH_RECORD"/>
    logger>

    <logger name="com.macro.mall" level="DEBUG">
        <appender-ref ref="LOG_STASH_BUSINESS"/>
    logger>
configuration>

配置要点解析

使用默认的日志配置

一般我们不需要自定义控制台输出,可以采用默认配置,具体配置参考console-appender.xml,该文件在spring-boot-${version}.jar下面。


<include resource="org/springframework/boot/logging/logback/defaults.xml"/>

<include resource="org/springframework/boot/logging/logback/console-appender.xml"/>

springProperty

该标签可以从SpringBoot的配置文件中获取配置属性,比如说在不同环境下我们的Logstash服务地址是不一样的,我们就可以把该地址定义在application.yml来使用。

例如在application-dev.yml中定义了这些属性:

logstash:
  host: localhost

logback-spring.xml中就可以直接这样使用:


<springProperty scope="context" name="APP_NAME" source="spring.application.name" defaultValue="springBoot"/>

<springProperty name="LOG_STASH_HOST" scope="context" source="logstash.host" defaultValue="localhost"/>

filter

在Logback中有两种不同的过滤器,用来过滤日志输出。

ThresholdFilter:临界值过滤器,过滤掉低于指定临界值的日志,比如下面的配置将过滤掉所有低于INFO级别的日志。

<filter class="ch.qos.logback.classic.filter.ThresholdFilter">
    <level>INFOlevel>
filter>

LevelFilter:级别过滤器,根据日志级别进行过滤,比如下面的配置将过滤掉所有非ERROR级别的日志。

<filter class="ch.qos.logback.classic.filter.LevelFilter">
    <level>ERRORlevel>
    <onMatch>ACCEPTonMatch>
    <onMismatch>DENYonMismatch>
filter>

appender

Appender可以用来控制日志的输出形式,主要有下面三种。

  • ConsoleAppender:控制日志输出到控制台的形式,比如在console-appender.xml中定义的默认控制台输出。

<appender name="CONSOLE" class="ch.qos.logback.core.ConsoleAppender">
 <encoder>
  <pattern>${CONSOLE_LOG_PATTERN}pattern>
 encoder>
appender>
  • RollingFileAppender:控制日志输出到文件的形式,可以控制日志文件生成策略,比如文件名称格式、超过多大重新生成文件以及删除超过多少天的文件。


<appender name="FILE_ERROR"
          class="ch.qos.logback.core.rolling.RollingFileAppender">

    <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
        
        <fileNamePattern>${LOG_FILE_PATH}/error/${APP_NAME}-%d{yyyy-MM-dd}-%i.logfileNamePattern>
        
        <maxFileSize>${LOG_FILE_MAX_SIZE:-10MB}maxFileSize>
        
        <maxHistory>${LOG_FILE_MAX_HISTORY:-30}maxHistory>
    rollingPolicy>
appender>
  • LogstashTcpSocketAppender:控制日志输出到Logstash的形式,可以用来配置Logstash的地址、访问策略以及日志的格式。


<appender name="LOG_STASH_ERROR" class="net.logstash.logback.appender.LogstashTcpSocketAppender">
    <destination>${LOG_STASH_HOST}:4561destination>
    <encoder charset="UTF-8" class="net.logstash.logback.encoder.LoggingEventCompositeJsonEncoder">
        <providers>
            <timestamp>
                <timeZone>Asia/ShanghaitimeZone>
            timestamp>
            
            <pattern>
                <pattern>
                    {
                    "project": "mall-tiny",
                    "level": "%level",
                    "service": "${APP_NAME:-}",
                    "pid": "${PID:-}",
                    "thread": "%thread",
                    "class": "%logger",
                    "message": "%message",
                    "stack_trace": "%exception{20}"
                    }
                pattern>
            pattern>
        providers>
    encoder>
    
    <connectionStrategy>
        <roundRobin>
            <connectionTTL>5 minutesconnectionTTL>
        roundRobin>
    connectionStrategy>
appender>

logger

只有配置到logger节点上的appender才会被使用,logger用于配置哪种条件下的日志被打印,root是一种特殊的appender,下面介绍下日志划分的条件。

  • 调试日志:所有的DEBUG级别以上日志;

  • 错误日志:所有的ERROR级别日志;

  • 业务日志:com.macro.mall包下的所有DEBUG级别以上日志;

  • 记录日志:com.macro.mall.tiny.component.WebLogAspect类下所有DEBUG级别以上日志,该类是统计接口访问信息的AOP切面类。

控制框架输出日志

还有一些使用框架内部的日志,DEBUG级别的日志对我们并没有啥用处,都可以设置为了INFO以上级别。


<logger name="org.slf4j" level="INFO"/>
<logger name="springfox" level="INFO"/>
<logger name="io.swagger" level="INFO"/>
<logger name="org.springframework" level="INFO"/>
<logger name="org.hibernate.validator" level="INFO"/>



Logstash配置详解

接下来我们需要配置下Logstash,让它可以分场景收集不同的日志,下面详细介绍下使用到的配置。

完全配置

input {
tcp {
mode => "server"
host => "0.0.0.0"
port => 4560
codec => json_lines
type => "debug"
}
tcp {
mode => "server"
host => "0.0.0.0"
port => 4561
codec => json_lines
type => "error"
}
tcp {
mode => "server"
host => "0.0.0.0"
port => 4562
codec => json_lines
type => "business"
}
tcp {
mode => "server"
host => "0.0.0.0"
port => 4563
codec => json_lines
type => "record"
}
}
filter{
if [type] == "record" {
mutate {
remove_field => "port"
remove_field => "host"
remove_field => "@version"
}
json {
source => "message"
remove_field => ["message"]
}
}
}
output {
elasticsearch {
hosts => ["es:9200"]
action => "index"
codec => json
index => "mall-tiny-%{type}-%{+YYYY.MM.dd}"
template_name => "mall-tiny"
}
}

配置要点

  • input:使用不同端口收集不同类型的日志,从4560~4563开启四个端口;

  • filter:对于记录类型的日志,直接将JSON格式的message转化到source中去,便于搜索查看;

  • output:按类型、时间自定义索引格式。



SpringBoot配置

在SpringBoot中的配置可以直接用来覆盖Logback中的配置,比如logging.level.root就可以覆盖节点中的level配置。

  • 开发环境配置:application-dev.yml

logstash:
  host: localhost
logging:
  level:
    root: debug
  • 测试环境配置:application-test.yml

logstash:
  host: 192.168.3.101
logging:
  level:
    root: debug
  • 生产环境配置:application-prod.yml

logstash:
  host: logstash-prod
logging:
  level:
    root: info



Kibana进阶使用

进过上面ELK环境的搭建和配置以后,我们的日志收集系统终于可以用起来了,下面介绍下在Kibana中的使用技巧!

  • 首先启动我们的测试Demo,然后通用调用接口(可以使用Swagger),产生一些日志信息;

7340ad51b3a5518a76a3c0cacbb317dc.webp
  • 调用完成后在Management->Kibana->Index Patterns中可以创建Index Patterns,Kibana服务访问地址:http://192.168.3.101:5601

d5c104cf0f434a9277a5c54a8214cec4.webp
  • 创建完成后可以在Discover中查看所有日志,调试日志只需直接查看mall-tiny-debug*模式的日志即可;

a3a32e59f8a2ee93c429c10bcd604b2a.webp
  • 对于日志搜索,kibana有非常强大的提示功能,可以通过搜索栏右侧的Options按钮打开;

8666bb8b6bbca72c1928577e39ca67f5.webp
  • 记录日志只需直接查看mall-tiny-record*模式的日志即可,如果我们想要搜索uri为/brand/listAll的记录日志,只需在搜索栏中输入uri : "/brand/listAll"

7764a48f4aa1017e73f4b340a7cee260.webp
  • 错误日志,只需直接查看mall-tiny-error*模式的日志即可;

6346fc309d3b393032fe5645ed7d5274.webp
  • 业务日志,只需直接查看mall-tiny-business*模式的日志即可,这里我们可以查看一些SQL日志的输出;

11d1beacd1e00d32c880cf7c4c1f6442.webp
  • 如果日志太大了,可以通过Elasticsearch->Index Management选择删除即可。

a99a182b10917239f4c8204ab688a249.webp



项目源码地址

https://github.com/macrozheng/mall-learning/tree/master/mall-tiny-log


有道无术,术可成;有术无道,止于术

欢迎大家关注Java之道公众号


好文章,我在看❤️

浏览 14
点赞
评论
收藏
分享

手机扫一扫分享

分享
举报
评论
图片
表情
推荐
点赞
评论
收藏
分享

手机扫一扫分享

分享
举报