[
https://issues.apache.org/jira/browse/NIFI-4071?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16889078#comment-16889078
]
Matt Burgess commented on NIFI-4071:
------------------------------------
As a follow-up, for Hive 3 the preferred approaches are PutHive3Streaming for
managed tables and PutORC/PutHDFS -> ReplaceText (with Hive DDL) -> PutHiveQL
for external tables. Both of these are more efficient and less error-prone than
trying to convert individual rows into HiveQL and then perform an INSERT.
> ConvertJSONToSQL does not support Hive
> --------------------------------------
>
> Key: NIFI-4071
> URL: https://issues.apache.org/jira/browse/NIFI-4071
> Project: Apache NiFi
> Issue Type: Improvement
> Components: Extensions
> Reporter: Matt Burgess
> Priority: Major
>
> Currently, the ConvertJSONToSQL processor does not support Hive as the target
> database, either using a HiveConnectionPool or a DBCPConnectionPool
> configured with a Hive driver. At the very least:
> 1) A SQLException occurs when determining the auto-increment value. This is
> due to a Hive bug (HIVE-13528
> [https://issues.apache.org/jira/browse/HIVE-13528]) where the column is not
> named according to the spec
> ([http://docs.oracle.com/javase/8/docs/api/java/sql/DatabaseMetaData.html#getColumns-java.lang.String-java.lang.String-java.lang.String-java.lang.String-]).
> 2) Column names are returned with the table name prepended (using a dot
> separator)
> 3) There may be other JDBC API calls that are not supported by the Hive JDBC
> driver.
> #1 could be solved by checking for "IS_AUTOINCREMENT" then failing over to
> "IS_AUTO_INCREMENT". #2 could be solved with a lastIndexOf("."), as is done
> in _org.apache.nifi.util.hive.HiveJdbcCommon#createSchema(java.sql.ResultSet,
> java.lang.String)_. #3 would take some investigation into the other JDBC API
> calls being made.
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
