Contact UsGet started

Creating an external HDFS data source

Written by
Updated at October 17, 2024

In Managed Service for Greenplum®, as an external data source with the HDFS connection type, you can use HDFS as part of Yandex Data Processing or other third-party HDFS services.

Create an external data source

  1. Go to the folder page and select Managed Service for Greenplum.
  2. Open the page of the Managed Service for Greenplum® cluster you need.
  3. In the left-hand panel, select  PXF.
  4. Click Create data source.
  5. Select the HDFS connection type.
  6. Enter a source name.
  7. Configure at least one optional setting.
  8. Click Create.

To add an HDFS data source to a Managed Service for Greenplum® cluster, use the create REST API method for the PXFDatasource resource or the PXFDatasourceService/Create gRPC API call and provide the following in the request:

  • Cluster ID in the clusterId parameter. To find out the cluster ID, get a list of clusters in the folder.
  • Source name in the name parameter.
  • External source settings in the hdfs parameter.

Sample REST API request

The example below shows how to create an external HDFS data source using the Managed Service for Greenplum® REST API. To create a source:

  1. Get an IAM token. It is used for authentication in the API.

  2. Add the IAM token to the following environment variable:

    export IAM_TOKEN=<token>

  3. Send a request using cURL:

    curl --location "https://mdb.api.cloud.yandex.net/managed-greenplum/v1/clusters/<cluster_ID>/pxf_datasources" \
    --header "Content-Type: text/plain" \
    --header "Authorization: Bearer ${IAM_TOKEN}" \
    --data "{
        \"datasource\": {
            \"name\": \"hdfs:csv\",
            \"hdfs\": {
                \"core\": {
                    \"defaultFs\": \"<storage_type:_DISK_or_ARCHIVE>\"
                }
            }
        }
    }"

    In the request body, specify the following parameters:

    • name: Source name, e.g., hdfs:csv.

    • defaultFs: Default data storage type (optional). The possible values include:

      • DISK: Data storage on a physical disk.
      • ARCHIVE: Archival data storage. In this case, you can store more data in HDFS, but their processing speed will be lower.

Greenplum® and Greenplum Database® are registered trademarks or trademarks of VMware, Inc. in the United States and/or other countries.

Previous
JDBC
Next
Hive