Elasticsearch fielddata unsupported while index mapping

Question

I'm trying to create an index in Elasticsearch through Python. I have a local instance of ES deployed and queries run fine. However, I have a schema. Here it is:

{
  "mappings": {
    "payment":{
      "properties":{
        "timestamp":{"type":"date"},
        "base_amount":{"type":"integer"},
        "derived_id":{"type":"keyword", "fielddata": true},
        "attempts":{"type":"integer"},
        "status":{"type":"text","fielddata":true},
        "error_code":{"type":"text","fielddata":true}
      }
    }
  }
}

Here is the code I am using to create this index

import json

import requests

schema = {
    "mappings": {
        "payment": {
            "properties": {
                "timestamp": {"type": "date"},
                "base_amount": {"type": "integer"},
                "derived_key": {"type": "keyword", "fielddata": True},
                "attempts": {"type": "integer"},
                "status": {"type": "text", "fielddata": True},
                "error_code": {"type": "text", "fielddata": True}
            }
        }
    }
}

index = 'http://localhost:9200/payment_index_2016_08_21'
r = requests.put(index, data=json.dumps(schema))

print r.content

The error I get is

{"error":{"root_cause":[{"type":"mapper_parsing_exception","reason":"Mapping definition for [derived_key] has unsupported parameters: [fielddata : true]"}],"type":"mapper_parsing_exception","reason":"Failed to parse mapping [payment]: Mapping definition for [derived_key] has unsupported parameters: [fielddata : true]","caused_by":{"type":"mapper_parsing_exception","reason":"Mapping definition for [derived_key] has unsupported parameters: [fielddata : true]"}},"status":400}

I don't understand why the fielddata = true is causing an issue, since I see it's allowed here https://www.elastic.co/guide/en/elasticsearch/reference/current/fielddata.html. Any clue what the issue behind this is?

Mohammad Mazraeh · Accepted Answer · 2017-06-05 18:58:09Z

3

You don't need to enable fielddata on keyword fields. you can do aggregations on keyword fields.

answered Jun 5, 2017 at 18:58

Mohammad Mazraeh

1,0847 silver badges12 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

gabbar0x Over a year ago

Yup it works. So what you mean is that on keywords, you can perform aggregation by default and we don't need to enable that by fielddata = true (since it's already enabled) and the purpose of fielddata is to perform aggreations, correct?

Mohammad Mazraeh Over a year ago

Yes are are correct. some tasks like aggregations, sorting, ... need different requirements than search and text fields are mean to be searched in by default an other fields can be aggregated/sorted on by default. So if you need sorting/aggregation/... on text fields you should enable fielddata for them which could be memory expensive!

gabbar0x Over a year ago

Done. Just changed it to text instead of keyword and it works. Thanks!

Mohammad Mazraeh Over a year ago

Good. You can take a look at https://www.elastic.co/guide/en/elasticsearch/reference/current/fielddata.html for more detailed information about fielddata.

Collectives™ on Stack Overflow

Elasticsearch fielddata unsupported while index mapping

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related