Reads records from a Hive table and converts each record into a StructuredRecord with the help of the specified schema (if provided) or the table's schema.


referenceName: This will be used to uniquely identify this source for lineage, annotating metadata, etc.

metastoreURI: The URI of Hive metastore in the format of thrift://<hostname>:<port>. Example: thrift://

tableName: The name of the Hive table. This table must exist.

databaseName: The name of the database. Defaults to 'default'.

partitions: Optional Hive expression filter for scan. This filter must only reference partition columns. Values from other columns will cause the pipeline to fail.

schema: Optional schema to use while reading from the Hive table. If no schema is provided, then the schema of the table will be used. Note: if you want to use a Hive table which has non-primitive types as a source, then you should provide a schema with all non-primitive fields dropped, otherwise your pipeline will fail.

  • CDAP Pipelines Plugin Type: batchsource
  • CDAP Pipelines Version: 1.8.5