Skip to main content
URL: http://<db.host>:<db.port>/insert/records/json

​
Overview

Kinetica provides a direct JSON ingest service to allow raw JSON to be translated into records and inserted into the database. If the target table exists, inserted records will be appended to it. If the target table does not exist, it will be created automatically, with type inferencing used to determine column types & sizes, before inserting data.
To ensure a particular column ordering with a created target table, use the columns_to_load parameter, specifying all columns in the desired order.
Either a single JSON object or multiple JSON objects (as an array) can be ingested at once: 
{
	"id": 14,
	"category": "Technology",
	"name": "Dot Matrix Printers",
	"description": "Impactful industrial implements for impressing ink iteratively",
	"stock": 27
}
For examples, see Examples or JSON Ingest Snippets.

​
Usage

The base JSON ingest endpoint URL is: 
http://<db.host>:9191/insert/records/json
The JSON payload should be passed as data in the request. Authentication credentials should be passed with the request, as well. Endpoint-specific options can be passed as request parameters. See /insert/records/frompayload under the option parameter for a complete list of supported options.
Both columns_to_load & columns_to_skip can only be used with named columns, not column positions, as JSON maps don’t lend themselves to positional references.

​
Examples

Below are templates of ingesting a JSON file of product data into a table, via the /insert/records/json REST endpoint: 
KINETICA_URL=http://localhost:9191
USERNAME=auser
PASSWORD=apassword
TABLE_NAME=product
JSON_FILE_NAME=products.json

curl -sS -X POST --header "Content-Type: application/json" \
--user "${USERNAME}:${PASSWORD}" \
-d @${JSON_FILE_NAME} \
${KINETICA_URL}/insert/records/json?table_name=${TABLE_NAME}
Endpoint-specific options can be passed as request parameters: 
KINETICA_URL=http://localhost:9191
USERNAME=auser
PASSWORD=apassword
OPTS_TABLE_NAME=product_options
JSON_FILE_NAME=products.json

# Quote the URL when passing multiple options, or the & will stop the URL
#   parsing and run the URL parsed up to that point as a background job
curl -sS -X POST --header "Content-Type: application/json" \
--user "${USERNAME}:${PASSWORD}" \
-d @${JSON_FILE_NAME} \
"${KINETICA_URL}/insert/records/json?table_name=${OPTS_TABLE_NAME}&batch_size=10000&columns_to_load=name,category,description&truncate_table=true"
Data can also be in-lined with the command: 
KINETICA_URL=http://localhost:9191
USERNAME=auser
PASSWORD=apassword
TABLE_NAME=product

curl -sS -X POST --header "Content-Type: application/json" \
--user "${USERNAME}:${PASSWORD}" \
-d '{
	"id": "15",
	"category": "Technology",
	"name": "DVDs",
	"description": "Lightweight storage for low-res screenplays and licensed software",
	"stock": "5000"
}' \
${KINETICA_URL}/insert/records/json?table_name=${TABLE_NAME}

​
Responses

The endpoint calls will respond with one of two message formats, one for successes and one for failures: 
{
	"status": "OK",
	"data_type": "insert_records_from_payload_response",
	"data":
	{
		"table_name": "<table name>",
		"type_id": "<type id>",
		"type_definition": "<type definition as json string",
		"type_label": "",
		"type_properties": <map of column name to extented properties list>,
		"count_inserted": <count of records inserted>,
		"count_skipped": <count of records not inserted>,
		"count_updated": <count of records upserted by primary key match>,
		"info": {}
	}
}
An example of each of these is as follows: 
{
    "status": "OK",
    "data_type": "insert_records_from_payload_response",
    "data": {
        "table_name": "test.product",
        "type_id": "15901623289898927486",
        "type_definition": "{\"type\":\"record\",\"name\":\"type_name\",\"fields\":[{\"name\":\"category\",\"type\":\"string\"},{\"name\":\"description\",\"type\":\"string\"},{\"name\":\"id\",\"type\":\"int\"},{\"name\":\"name\",\"type\":\"string\"},{\"name\":\"stock\",\"type\":\"long\"}]}",
        "type_label": "test.product",
        "type_properties": {
            "category": [
                "data",
                "char64"
            ],
            "description": [
                "data"
            ],
            "id": [
                "data"
            ],
            "name": [
                "data",
                "char128"
            ],
            "stock": [
                "data"
            ]
        },
        "count_inserted": 1,
        "count_skipped": 0,
        "count_updated": 0,
        "info": {
            "distribute_all": "true",
            "external_file_info": "{'column_names':'[\\'category\\',\\'description\\',\\'id\\',\\'name\\',\\'stock\\']','has_kinetica_header':'0','has_name_header':'1','header':'[]','inferred_batch_size':'0','inferred_num_records':'14','is_header_inferred':'0','is_text_column_delimiter_inferred':'0','is_type_from_kinetica_header':'0','is_type_inferred':'1','text_column_delimiter':'44'}",
            "field_descs": "{'category':{'card_est':'3','count':'14','dict_encoded':'0','field_type':'string','max_string_length':'15','nullable':'0'},'description':{'card_est':'14','count':'14','dict_encoded':'0','field_type':'string','max_string_length':'85','nullable':'0'},'id':{'card_est':'14','count':'14','dict_encoded':'0','field_type':'unsigned','max':'14','max_string_length':'2','nullable':'0'},'name':{'card_est':'14','count':'14','dict_encoded':'0','field_type':'string','max_string_length':'19','nullable':'0'},'stock':{'card_est':'14','count':'14','dict_encoded':'0','field_type':'unsigned','max':'1000000','max_string_length':'7','nullable':'0'}}"
        }
    }
}