(?m) 标记

<pre name="code" class="html">在和 codec/multiline 搭配使用的时候，需要注意一个问题，grok 正则和普通正则一样，默认是不支持匹配回车换行的。就像你需要 =~ //m 一样也需要单独指定，具体写法是在表达式开始位置加 (?m) 标记。如下所示：

match => {
    "message" => "(?m)\s+(?<request_time>\d+(?:\.\d+)?)\s+"
}
input {
        file {
                type => "zj_mysql"
                path => ["/data01/applog_backup/zjzc_log/zj-mysql01-slowlog.*"]
                 codec => multiline {
      pattern => "^\s+#\s+User@Host:"
      negate => true
      what => "previous"
    }

        }

       file {
                type => "wj_mysql"
                path => ["/data01/applog_backup/winfae_log/wj-mysql01-slowlog.*"]
                 codec => multiline {
      pattern => "^\s+#\s+User@Host:"
      negate => true
      what => "previous"
    }

        } 

    }

filter {
  # drop sleep events
  grok {
    match => { "message" => "SELECT SLEEP" }
    add_tag => [ "sleep_drop" ]
    tag_on_failure => [] # prevent default _grokparsefailure tag on real records
  }
  if "sleep_drop" in [tags] {
    drop {}
  }
  grok {
    match => [ "message","(?m)\s*# User@Host:\s+\S+\[%{USER:user}\]\s+@\s+\[%{IP:clientip}\]\s+(?<id>(\S+\s+)*\S+)\s*#\s+Query_time:\s+%{NUMBER:Query_time}\s+Lock_time: %{NUMBER:lock_time}\s+Rows_sent: %{NUMBER:rows_sent}\s+Rows_examined: %{NUMBER:rows_examined}\s*
\s*SET\s+timestamp=%{NUMBER:timestamp};\s*(?<query>(\s*\S+\s*).*)\s*"
]
  }
  date {
    match => [ "timestamp", "UNIX" ]
    remove_field => [ "timestamp" ]
  }
}
output {
     if [type] == "zj_mysql" {
        redis {
                host => "192.168.32.67"
                data_type => "list"
                key => "zj_mysql:redis"
                port=>"6379"
                password => "1234567"
        }
}
      else if [type] == "wj_mysql"{
       redis {
                host => "192.168.32.67"
                data_type => "list"
                key => "wj_mysql:redis"
                port=>"6379"
                password => "1234567"
        }
}
}