Properties reference: HBase connector

This topic lists all properties that you can set to configure the stage.

Connection

For more information, see the Connecting to a HBase data source topic.

Hadoop identity
Something that uniquely identifies Hadoop cluster. It may be a name of your choosing or e.g. clusterId
  • Type: string
HBase identity
Something that uniquely identifies HBase cluster. It may be a name of your choosing or e.g. clusterId
  • Type: string
Input method for XML files
Choose the method to provide the core-site and hbase-site xml files.
  • Type: selection
  • Default: File path
  • Values:
    • File path
    • File content
hbase-site.xml
Location of hbase-site.xml in node's local filesystem
  • Type: string
  • Default: /home/dsadm/hbase/hbase-site.xml
core-site.xml
Location of core-site.xml in node's local filesystem
  • Type: string
  • Default: /home/dsadm/hbase/core-site.xml
HBase client jars
Provide the path (without wildcards) to the directory containing HBase client jars or list of jars separated by semicolon
  • Type: string
  • Default: /home/dsadm/hbase-client
Simple authentication user name
Simple authentication user name
  • Type: string
  • Default: dsadm
Use Kerberos
Use Kerberos
  • Type: boolean
  • Default: false
krb5.conf location
path of krb5.conf file
  • Type: string
  • Default: /etc/krb5.conf
Principal
Principal
  • Type: string
  • Default: user@REALM
Use Keytab
Use Keytab
  • Type: boolean
  • Default: false
Keytab
Path to keytab
  • Type: string
Password
Password
  • Type: protected string
  • Default: password
Use ticket cache
Use ticket cache
  • Type: boolean
  • Default: false

Usage

HBase Namespace
HBase Namespace
  • Type: string
  • Default: default
Target table
Name of a target table
  • Type: string
  • Default: table
Strict type checking
If strict type checking is enabled exception is thrown when value of given type can be parsed but there would be some bytes left. When disabled only warning is logged and also if reading nullable data field fails, its value will be set to null instead of throwing an Exception
  • Type: boolean
  • Default: true
Use parallel read
Split reading data to all available nodes to speed up the process
  • Type: boolean
  • Default: true
Type of row keys in the target table
The split algorithm chosen for the table is based on the type of row keys
  • Type: selection
  • Default: Uniform byte arrays
  • Values:
    • Uniform byte arrays
    • Numeric strings with values greater than zero
    • Hexadecimal strings with values greater than zero
Detect regions modifications (splits & merges)
Detect splits & merges that can impact read performance
  • Type: category
Enable detecting regions changes
Detect splits & merges that can impact read performance
  • Type: boolean
  • Default: true
Detect after reading every
Detection will be run after reading a specified number of rows
  • Type: integer
  • Default: 100000
Write mode
Write mode
  • Type: selection
  • Default: Put
  • Values:
    • Put
    • Delete row
    • Delete qualifiers
    • Append qualifiers values
For more information, see the Configuring the HBase connector for partitioned write topic.
Autoflush enabled
Enables flushing the buffer after each row. Auto flush is disabled once input link has corresponding reject link.
  • Type: boolean
  • Default: true
Java settings
Specify the customized Java settings for the stage.
  • Type: category
JVM options
Specify the additional Java Virtual Machine options.
  • Type: string