Why the slow query doesn't log in slowlog file

Elastic Stack Elasticsearch
1

Server Version: 6.8.6

On the client side, the response indicates that the query execution time was 394ms:

{"took":394,"timed_out":false,"_shards":{"total":5,"successful":5,"failed":0},"hits":

However, despite setting the following configuration:

{
   "index": {
      "search": {
         "slowlog": {
            "level": "debug",
            "threshold": {
               "fetch": {
                  "debug": "300ms"
               },
               "query": {
                  "debug": "300ms"
               }
            }
         }
      }
   }
}

No slow log pertaining to this query can be found. It seems that the slow log on the server side only captures the time spent on a single node and a single shard. If this assumption is correct, how can I configure the server to log all slow queries whose total execution time exceeds the specified threshold?