Filebeat、Logstash读取多个topic并输出到ES集群

一、配置filebeat.yml

filebeat.inputs:
- type: log
  enabled: true
  paths:
    - /usr/local/nginx/logs/access.log
#此处加上fields和log_topics字段，log_topics=nginx和下面的log_topics=httpd主要用于区分两个日志，这两个字段的值可以自定义
  fields:
    log_topics: nginx
  #最好在每个日志里都写上下面这三行JSON的配置，不然日志写到kafka，在输出到ES的时候，会以整条message的形式显示出来，不利于过滤，亲测过。
  json.keys_under_root: true
  json.overwrite_keys: true
  json.add_error_key: true

- type: log
  enabled: true
  paths:
    - /var/log/httpd/access_log.ls_json
  fields:
    log_topics: httpd
  json.keys_under_root: true
  json.overwrite_keys: true
  json.add_error_key: true

#下面是output.kafka的配置
#=========================== Kafka Output=====================
output.kafka:
  enabled: true
  hosts: ["10.1.1.17:9092"]
  #此处topic取值是上面新加的字段的值
  topic: '%{[fields][log_topics]}'

配置完成后，重启filebeat服务，没有报错的话，可以在kafka中创建了nginx和httpd两个topic：

[root@elk kafka]# bin/kafka-topics.sh  --zookeeper 10.1.1.17:2181 --list
__consumer_offsets
httpd
nginx

topic创建成功了，下面接着配置logstash：

二、配置Logstash

[root@elk kafka]# cat /etc/logstash/conf.d/kafka-es.conf 
input {
    kafka  {
        codec => "json"
        #此处把两个topic的名称都写进去，用英文状态的逗号隔开
        topics => ["nginx","httpd"]
        bootstrap_servers => "10.1.1.17:9092"
        auto_offset_reset => "latest"
   }
}

filter {
#filter部分也要加个if判断，不同的topic采取不同的处理措施
#topic=nginx
    if[fields][log_topics] == "nginx" {
        date {
            match => ["requesttime", "dd/MMM/yyyy:HH:mm:ss Z +08:00"]
            target => "@timestamp"
        }
        ruby {
            code => "event.set('timestamp', event.get('@timestamp').time.utc+8*60*60)"
        }
        mutate {
            convert => ["timestamp", "string"]
            gsub => ["timestamp", "T([\S\s]*?)Z", ""]
            gsub => ["timestamp", "-", "."]
        }
        mutate {
            remove_field => ["_index","_id","_type","_version","_score","host","log","referer","input","path","agent"]
        }
    }
#topic=httpd
    if[fields][log_topics] == "httpd" {
        date {
            match => ["requesttime", "dd/MMM/yyyy:HH:mm:ss Z +08:00"]
            target => "@timestamp"
        }
        ruby {
            code => "event.set('timestamp', event.get('@timestamp').time.utc+8*60*60)"
        }
        mutate {
            convert => ["timestamp", "string"]
            gsub => ["timestamp", "T([\S\s]*?)Z", ""]
            gsub => ["timestamp", "-", "."]
        }
        mutate {
            remove_field => ["_index","_id","_type","_version","_score","host","log","referer","input","path","agent"]
        }
    }
}

output {
    elasticsearch {
        hosts => ["10.1.1.17:9200"]
        #此处的索引名称，我们设定为引用filebeat中配置的字段，这样就会自动创建两个带日期的索引，
        index =>  '%{[fields][log_topics]}+%{timestamp}'
    }
} 

三、Kibana查看索引

然后启动logstash，无报错的话，即可在kibana的管理界面-索引管理看到新建的nginx+2020.03.11和httpd+2020.03.11索引：我们据此创建索引模式，分别为nginx+2020.03.11和httpd+2020.03.11，然后分别查看日志情况：
可以看到http的日志，都在httpd索引下，nginx的日志，都在nginx索引下：

至此，配置完毕！