filebeat日志收集

以nginx错误日志为例，演示日志处理流程

filebeat--logstash--es

filebeat--kafka--logstash--es

#filebeat使用systemd管理
/usr/lib/systemd/system/filebeat.service 
[Unit]
Description=Filebeat
Documentation=http://www.elastic.co
Wants=network-online.target
After=network-online.target
[Service]
ExecStart=/usr/local/filebeat/filebeat -c /usr/local/filebeat/filebeat.yml
Restart=always
[Install]
WantedBy=multi-user.target

#logstash使用systemd管理
#如果有多个logstash配置文件,可以使用-f指定目录
/usr/lib/systemd/system/logstash.service 
[Unit]
Description=logstash
Documentation=http://www.elastic.co
Wants=network-online.target
After=network-online.target
[Service]
Environment=JAVA_HOME=/usr/java/jdk1.8.0_211
ExecStart=/usr/local/logstash/bin/logstash -f /usr/local/logstash/config/logstash.conf -l /usr/local/logstash/logs
Restart=always
[Install]
WantedBy=multi-user.target

#启动nginx容器，映射日志目录
docker run -d --name=nginx --net=host -v /tmp/nginx_log:/var/log/nginx nginx

#nginx错误日志：

2019/09/21 17:00:08 [error] 7#7: *9 open() "/usr/share/nginx/html/api" failed (2: No such file or directory), client: 192.168.3.102, server: localhost, request: "GET /api HTTP/1.1", host: "192.168.3.100"

• filebeat--logstash--es示例

#filebeat输出logstash示例
/usr/local/filebeat/filebeat.yml 
filebeat.inputs:
- type: log
  paths:
    - /tmp/nginx_log/error.log
  
  multiline.pattern: ^\d{4}/\d{2}/\d{2}\s\d{2}:\d{2}:\d{2}
  #匹配nginx日志时间格式  2019/09/21 17:00:08
  multiline.negate: true
  multiline.match: after
  exclude_files: [".gz$"]
  tail_files: true
  
  #增加输出字段，tags为数组形式，fields.id为键值对形式
  tags: ["nginx-100"]
  fields:
    id: "nginx-100"
output.logstash:
  hosts: ["192.168.3.100:5044","192.168.3.101:5044"]
  loadbalance: true

#输出到单个logstash
#output.logstash:
#  hosts: ["127.0.0.1:5044"]

#logstash输出到es示例；根据fileds.id来划分索引
/usr/local/logstash/config/logstash.conf 
input {
  beats {
    port => 5044
  }
}
output {
  elasticsearch {
    hosts => ["http://192.168.3.100:9200","http://192.168.3.101:9200","http://192.168.3.102:9200"]
    index => "%{[fields][id]}-%{+YYYY.MM.dd}"
    user => "elastic"
    password => "HkqZIHZsuXSv6B5OwqJ7"
  }
}

• filebeat--kafka--logstash--es示例

#filebeat输出到kafka示例
/usr/local/filebeat/filebeat.yml 
filebeat.inputs:
- type: log
  paths:
    - /tmp/nginx_log/error.log
  
  multiline.pattern: ^\d{4}/\d{2}/\d{2}\s\d{2}:\d{2}:\d{2}
  #匹配nginx日志时间格式  2019/09/21 17:00:08
  multiline.negate: true
  multiline.match: after
  exclude_files: [".gz$"]
  tail_files: true
  
  #增加输出字段，tags为数组形式，fields.id为键值对形式
  tags: ["nginx-kafka-100"]
  fields:
    id: "nginx-kafka-100"
output.kafka:
  hosts: ["192.168.3.100:9092", "192.168.3.101:9092", "192.168.3.102:9092"]

  topic: '%{[fields.id]}'
  partition.round_robin:
    reachable_only: false

  required_acks: 1
  compression: gzip
  max_message_bytes: 1000000

#kafka输出到es示例
/usr/local/logstash/config/logstash.conf
input {
    kafka {
      group_id => "logstash"
      topics => ["nginx-kafka-100"]
      bootstrap_servers => "192.168.3.100:9092,192.168.3.101:9092,192.168.3.102:9092"
      consumer_threads => "1"
      fetch_max_bytes => "26214400"
      codec => plain
  }
}
filter {
  json {
    source => "message"
  }
}
output {
  elasticsearch {
    hosts => ["http://192.168.3.100:9200","http://192.168.3.101:9200","http://192.168.3.102:9200"]
    index => "%{[fields][id]}-%{+YYYY.MM.dd}"
    user => "elastic"
    password => "HkqZIHZsuXSv6B5OwqJ7"
  }
}

参考：

https://www.elastic.co/guide/en/beats/filebeat/current/kafka-output.html

https://www.elastic.co/guide/en/beats/filebeat/current/logstash-output.html

https://www.elastic.co/guide/en/logstash/current/plugins-filters-json.html