GuidesAPI ReferenceDiscussions
GuidesBlogPlatform

HDFS File System

How to set up Databand environment for Hadoop Distributed File System.

Before You Begin

You must have the dbnd-hdfs plugin installed.

To Set up an Environment for Hadoop Distributed File System

  1. To set up HDFS access, in the HdfsConfig object or [hdfs] section of the project.cfg file, define the following properties:
  • Host (localhost by default)
  • Port (50070 by default)
  • User (root by default)
  • Client Type (knox - by default, allowed values - insecure, kerberos or knox).

Here is the configuration example:

[hdfs]
host = localhost
port = 50070
user = root
client_type = knox
knox_gateway=bigdata
  1. Depending on your client type, additionally define the following properties:
knox_base = parameter.c(default=None)[str]
knox_secure = parameter.c(default=False)[bool]
knox_gateway = parameter.c(default=None)[str]
knox_password = parameter.c(default=None)[str]
knox_cookies = parameter.c(default=None)[str]
knox_bearer_token = parameter.c(default=None)[str]
knox_bearer_token_encode = parameter.c(default=True)[bool]
kerberos_mutual_authentication = parameter.c(default=KerberosMutualAuthentication.REQUIRED)[str]
kerberos_service = parameter.c(default="HTTP")[str]
kerberos_delegate = parameter.c(default=False)[bool]
kerberos_force_preemptive = parameter.c(default=False)[bool]
kerberos_principal = parameter.c(default=None)[str]
kerberos_hostname_override = parameter.c(default=None)[str]
kerberos_sanitize_mutual_error_response = parameter.c(default=True)[bool]
kerberos_send_cbt = parameter.c(default=True)[bool]
  1. To write outputs to HDFS, specify the output configuration:
[output]
path_task = hdfs://{task_target_date}{sep}{task_name}{sep}{task_name}{task_class_version}_{task_signature}{sep}{output_name}{output_ext}

What’s Next