需要帮助从流利的文件日志中删除\ n

Question

我正在使用流利的和winston将所有日志记录到文件和elasticsearch中。我正在从服务器记录JSON数据。以下是登录文件的内容。

2018-06-06T15:54:37+05:30   service.api {"level":"infp","message":"{ user: null,\n  serviceid: null,\n  path: null,\n  method: null,\n  agent: null,\n  code: null,\n  clientip: null,\n  size: null,\n  message: 'API LOGGER INITIATED',\n  http_x_forwarded_for: null,\n  parameters: null,\n  query: null,\n  body: null,\n  controller: null,\n  controller_method: null,\n  correlation_id: null,\n  fileInfo: null,\n  time: 1528280677002 }","meta":{"end":true}}

我需要从文件中删除\ n并将其格式化为JSON而不是字符串。还想将字段“message”重命名为“log”。

以下是我精通的配置

## SYSTEM CONFIGURATION
<system>
  <log>
    format json
    time_format %Y-%m-%d
  </log>
</system>

########################################################
#### External plugins ##################################
#### 1. fluent-plugin-forest ###########################
#### https://github.com/tagomoris/fluent-plugin-forest##
########################################################

## INFO LOGS
<source>
  @type forward
  bind 127.0.0.1
  port 8001
</source>



<match service.**>
  @type copy
  <store>
    ## writing logs to file as per microservice
    @type forest 
    subtype file
    <template>
      append true
      compress gzip
      <format>
        @type json
        add_newline true  // tried with both true/false
        message_key record  true  // Not working
      </format>
      <buffer time>
        # @type file
        timekey         1
        timekey_wait    1
        retry_max_times 5
      </buffer>
      path /var/log/fluentd/__TAG__/log/file
    </template>
  </store>
  <store>
    @type grep
    key level
    # pattern verbose
    add_tag_prefix access
  </store>
</match>

## Writing logs to elastic search

<match access.**>
  @type elasticsearch
  host localhost
  port 9200
  logstash_format true
</match>