We are just exploring KNIME Big Data connectors for the first time under 30-day trial, trying to connect to Cloudera sandbox (Quickstarts for CDH 5.10 running under Docker). In our first test, we can establish a connection from KNIME 3.3 using Hive Connector successfully. But Database Table Connector gives this error:
Database Table Selector 2:6 Error while validating SQL query 'SELECT * FROM customers' : Unregonized Thrift TTypeId value: STRUCT_TYPE
Can you confirm if this is likely a configuration problem at our end, or if KNIME Big Data extensions are not compatible with CDH 5.10? If the latter, what is the most recent Cloudera sandbox we should be using?
this sounds like an issue with the builtin JDBC driver for Hive we are shipping as part of KNIME Big Data Connectors, which will be updated in the near-term. In the meantime you can try and use the proprietary Hive JDBC driver from Cloudera:
For Hive this will give you a file called
hive_jdbc_184.108.40.2060.zip (or similar)
Unzip this file. Inside it you find another zip file:
Cloudera_HiveJDBC41_2.5.18.zip (or similar)
If you unzip that one you get a folder with the driver jars. Add this folder in KNIME AP under File > Preferences > KNIME > Databases (via Directory).
Then restart KNIME AP and rerun the Hive query.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.