Redundant should block or query_string in query?

Elastic Stack Elasticsearch
1

hey there,
I can get the same result using two different approach for a specific query.
The first one uses several should terms

"should": [
            {
              "terms": {
                   "fieldname": ["value1"]
               }
              },
               ...
               ...
               ... 
               {
               "terms": {
                   "fieldname": ["valueX"]
               }
              }
            ],
             "minimum_should_match": "80%"

while the second one will use the block:

{
      "query_string" : {
         "default_field" : "fieldname",
         "query" : "value1 OR value2 ......OR valueX",
         "minimum_should_match": "80%"
         }
 }

is there any suggestion to use one of them and to avoid any load or to increase speed?
I don't know exactly how should directive works so I don't if it will query my index for every block or just one time.
Currently, the first scenario is in place and queries are taking more or less 500ms. Achieving a time reduction even just 100ms could be important but I have not a chance to reproduce the same load in the second scenario so I have to start from the theory.

2

Is there anyone that can try to address me?

3

Internally those will likely be rewritten to similar queries, so I would use the one I feel more comfortable with. You can use the explain field to compare the two queries with each other.

Hope that helps!