0

Objective: load csv to a hive/impala/hdfs table Tools: Knime Scenario: I am using table creator ( just 1 row with 1 value), HIve connector( this connects to host without issues), HDFS connection and hive loader. When I configure HDFS connection node with hive loader with host and port info same as hive connector I get an error: Execute failed: End of File Exception between local host is: "Cyuyuyu/xxx.xxx.x.xxx"; destination host is: "xxxnode04.xx.xx.com":21050; : java.io.EOFException; For more details see: http://wiki.apache.org/hadoop/EOFException.

What am I doing wrong? Thanks in advance.

DrSD
  • 151
  • 2
  • 12

1 Answers1

0

When I configure HDFS connection node with hive loader with host and port info same as hive connector

The port for HDFS and Hive are not the same. Check on which port HDFS is configured on your destination machine and use this one.

Mareike Hoeger
  • 61
  • 6
  • Thanks for replying and the help. Yes, I tried this but got an error: Connection to hdfs://user@xxx.xx.xx.com:8020 failed SIMPLE authentication is not enabled. Available:[TOKEN,KERBEROS] When I used KERBEROS: Connection to hdfs:///user@xxx.xx.xx.com:8020 failed failure to login using ticket cache file null – DrSD Jul 10 '18 at 17:37
  • Do i need to configure Kerberos client to achieve this? – DrSD Jul 10 '18 at 17:39
  • Yes if you are connecting to a Kerberos secure cluster you have to configure your kerbveros client and get e ticket before you run the connection node. – Mareike Hoeger Jul 26 '18 at 07:34