关于我之前的问题：如何在elasticsearch上只添加新文档或更改的文档？
对于相同的场景，现在，我使用python logstash async将json对象/字典插入日志索引。
示例代码：

test_logger = logging.getLogger('python-logstash-logger')
test_logger.setLevel(logging.INFO)
async_handler = AsynchronousLogstashHandler(host, port, database_path=None)
test_logger.addHandler(async_handler)
extra ={
          'record_id':'102',
          'status':'started'
       }
test_logger.info('python-logstash: records', extra=extra)

回答如下：https://stackoverflow.com/a/63588420/4091956 （适用于不使用logstash的实现）对于我前面的问题，如何使 record_id 在额外参数中传递文档id，以便不插入重复记录？它当前显示为附加字段内的子字段，即源字段内的子字段。

{
  "_index": "logstash-2020.08.25",
  "_type": "_doc",
  "_id": "xy121",
  "_version": 1,
  "_score": null,
  "_source": {
    "program": "main.py",
    "@timestamp": "2020-08-25T20:11:01.065Z",
    "@version": "1",
    "logsource": "",
    "pid": 13864,
    "host": "",
    "level": "INFO",
    "port": 61210,
    "type": "python-logstash",
    "extra": {
      "record_id": "88943",
      "status":"started" 
    },
    "message": "python-logstash: records"
  },
  "fields": {
    "@timestamp": [
      "2020-08-25T20:11:01.065Z"
    ]
  },
  "sort": [
    1690386261065
  ]
}

这里的目标是确保不通过logstash添加重复的记录。