Logstash Grok误差

Question

My logstash configuration is giving me this error:

每当我运行以下命令时: /opt/logstash/bin/logstash -f /etc/logstash/conf.d/logstash.conf -auto-调试

reason=>"Expected one of #, {, ,, ] at line 27, column 95 (byte 677) after filter {\n\n\tif [type] == \"s3\" {\n\t\tgrok {\n\t\n  \t\t\tmatch => [\"message\", \"%{IP:client} %{USERNAME} %{USERNAME} \\[%{HTTPDATE:timestamp}\\] (?:\"", :level=>:error, :file=>"logstash/agent.rb", :line=>"430", :method=>"create_pipeline"}

这与我在Grok调试器中检查的pattern.But有关，它为我提供了所需的answer.Please帮助。

Here is my logstash configuration: 

input {
    s3 {
      access_key_id => ""
      bucket => ""
      region => ""
      secret_access_key => ""
      prefix => "access"
      type => "s3"
      add_field => { source => gzfiles }
      sincedb_path => "/dev/null"
      #path => "/home/shubham/logstash.json"
      #temporary_directory => "/home/shubham/S3_temp/"

      backup_add_prefix => "logstash-backup"
      backup_to_bucket => "logstash-nginx-overcart"


}

}

 filter {

        if [type] == "s3" {
                grok {

                        match => ["message", "%{IP:client} %{USERNAME} %{USERNAME} \[%{HTTPDATE:timestamp}\] (?:"%{WORD:request} 
        %{URIPATHPARAM:path} HTTP/%{NUMBER:version}" %{NUMBER:reponse} %{NUMBER:bytes} "%{USERNAME}" %{GREEDYDATA:responseMessage})"]
        }
        }
        }

 output {
    elasticsearch {
    hosts => ''
    index => "accesslogs"
}
  }

Answer 1

在您的匹配分配中有几个未转义的“字符”(例如，在用户名var周围)，这些字符跳过解析器。如果您用\\转义那些字符，它应该可以工作。