Self-managed installation and maintenance > Installation and configuration guide > DataRobot advanced configuration section > Data ingestion from Hive with Kerberos Authentication

Data ingestion from Hive with Kerberos Authentication¶

On-premise installations of DataRobot support data ingestion from Hive metastore with Kerberos authentication using a JDBC driver.

Prerequisites¶

Hive JDBC driver .jar files
Kerberos principal and keytab file
krb5.conf file from the target Hive/Hadoop cluster
TLS truststore .jks file to use accessing Hive with TLS
Network connectivity from DR cluster to KDC (Kerberos key distribution center) server

Configuration¶

Set up client secrets from krb5.conf, keytab file, and truststore .jks file:

kubectl -n DR_CORE_NAMESPACE create secret generic hive-configuration --from-file=krb5.conf=./krb5.conf --from-file=datarobot.keytab=./datarobot.keytab --from-file=cacerts.jks=./cacerts.jks

Modify your values.yaml:

core:
  config_env_vars:
    # kerberos related settings
    KERBEROS_ENABLE: "True"
    ENABLE_DIRECT_KERBEROS_AUTHENTICATION: "True"
    KRB5_CONFIG: "/etc/krb5.conf"
    KERBEROS_KEYTAB_FILE: "/etc/datarobot.keytab"
    KERBEROS_PRINCIPAL: "admin/cluster1@EXAMPLE.COM"
    KERBEROS_INIT_ENABLE: "False"

    # mounting the kerberos items
    DSS_EXTRA_SECRET_MOUNTS: '[{"secret_name": "hive-configuration", "mount_path": "/etc/krb5.conf", "sub_path": "krb5.conf"},{"secret_name": "hive-configuration", "mount_path": "/etc/datarobot.keytab", "sub_path": "datarobot.keytab"},{"secret_name": "hive-configuration", "mount_path": "/opt/datarobot/etc/hadoop/cacerts.jks", "sub_path": "cacerts.jks"}]'

# -- Parameters for the DataRobot `datasets-service` sub-chart {: #parameters-for-the-datarobot-datasets-service-sub-chart }
datasets-service:
  component:
    api:
      # -- Set number of replicas
      replicaCount: 1
  extraVolumes:
  - name: dss-extra-secret-mounts
    secret:
      secretName: hive-configuration
      defaultMode: 420
  extraVolumeMounts:
  - name: dss-extra-secret-mounts
    mountPath: /etc/krb5.conf
    subPath: krb5.conf
    readOnly: true
  - name: dss-extra-secret-mounts
    mountPath: /etc/datarobot.keytab
    subPath: datarobot.keytab
    readOnly: true
  - name: dss-extra-secret-mounts
    mountPath: /opt/datarobot/etc/hadoop/cacerts.jks
    subPath: cacerts.jks
    readOnly: true

JDBC driver creating¶

After installing DataRobot with the given values, in the DataRobot UI, open User Settings and select Data Connections > Drivers.

Click + Add new driver, and then create a driver with the following settings:

Configuration: Custom
Class name: org.apache.hive.jdbc.HiveDriver
Driver file(s): Hive jdbc jar files

Configure New Data Connection¶

Navigate to the Data Connections tab and click + Add new data connection.

Click My Drivers.
Select the Hive driver from the dropdown.
JDBC URL:

jdbc:hive2://<host>:10000/default;principal=admin/cluster1@EXAMPLE.COM;ssl=true;sslTrustStore=/opt/datarobot/etc/hadoop/cacerts.jks;trustStorePassword=changeit

Click Add New Connection.
Click Test Connection. A Test Data Connection window will appear.

Note: The username and password fields can contain any data as that information is not passed to Hive.