[BUG] aggregation on nested field returns incorrect result

[LantaoJin]

Describe the bug

Aggregation value_count on field of nested type returns incorrect result. This issue only happens on nested type, object type without nested looks fine.

Related component

Search:Aggregations

To Reproduce

Repro:

# create an test index with nested field
PUT test_nested_field_index
{
  "mappings" : {
    "properties": {
      "patients": { 
        "type" : "nested"
      }
    }
  }
}
# add some testing data
PUT test_nested_field_index/_doc/100
{ 
  "patients": [ 
    {"name" : "John Doe", "age" : 56, "smoker" : true},
    {"name" : "Mary Major", "age" : 85, "smoker" : false}
  ] 
}
# value_count should return 2 instead of 0
GET test_nested_field_index/_search
{
  "size": 0,
  "aggs": {
    "count_patients_name": {
      "value_count": {
        "field": "patients.name.keyword"
      }
    }
  }
}

Expected behavior

value_count should return 2 instead of 0

Additional Details

Plugins Please list all plugins currently enabled.

Screenshots If applicable, add screenshots to help explain your problem.

Host/Environment (please complete the following information):

• OS: Mac
• Version 2.14

Additional context Add any other context about the problem here.

[mch2]

possibly related - https://github.com/opensearch-project/OpenSearch/pull/13324 / https://github.com/opensearch-project/OpenSearch/pull/14208 / https://github.com/opensearch-project/OpenSearch/issues/13288

[jed326]

@LantaoJin

Please correct me if my understanding here is wrong, but it looks like your agg is on the field patients.name.keyword but from the document it should just be patients.name.

Furthermore, when I try to do your value_count agg on just patients.name, I get the following:

"root_cause": [
            {
                "type": "illegal_argument_exception",
                "reason": "Text fields are not optimised for operations that require per-document field data like aggregations and sorting, so these operations are disabled by default. Please use a keyword field instead. Alternatively, set fielddata=true on [patients.name] in order to load field data by uninverting the inverted index. Note that this can use significant memory."
            }
        ]

Given that I don't think there are any problems here then, could you please double check?

[jed326]

Also, I think in order to interact with nested fields in an aggregation, a nested aggregation must be used. See: https://opensearch.org/docs/latest/aggregations/bucket/nested/

[kkewwei]

@jed326, @mch2, I test this, it'ok, it seems not a problem.

GET test_nested_field_index/_search
{
   "aggs": {
      "patients1": {
         "nested": {
            "path": "patients"
         },
         "aggs": {
            "aggs1": {
               "value_count": {
                  "field": "patients.name.keyword"
               }
            }
         }
      }
   }
}

[jed326]

Thanks @kkewwei! I also took at look at this, I think the problem is the nested aggregation was not used in the original query. I'm going to close this issue for now, @LantaoJin please re-open if there's any other issues here, thanks!