DS8 TD Connector and Pimary Indexes
Posted: Fri Apr 23, 2010 12:46 pm
I am trying to build a generic job to load a dataset to a target table in Teradata. We we are using the Teradata Connector. The job is generic, so I want the job to be abel to create/replace a target table if needed. The problem is that I cannot figure out how to specify a primary index.
By default, Teradata will use the first field in a table's DDL, or in the case of DataSTage, the first field of a record to define the Primary Index. The PI is used to spread your data across the amps to allow for parallel processing. If that first field has low cardinality, then you will get data skew - some amps having more data than others.
So, is there a way to specify what field or fields you want to use for a Primary Index in Teradata connector? And if so, can it be parameterized? If memory serves me correctly, you could do this with the old Teradata Enterprise stage that used FastLoad, but we need to use Connector.
Thanks!
Brad
By default, Teradata will use the first field in a table's DDL, or in the case of DataSTage, the first field of a record to define the Primary Index. The PI is used to spread your data across the amps to allow for parallel processing. If that first field has low cardinality, then you will get data skew - some amps having more data than others.
So, is there a way to specify what field or fields you want to use for a Primary Index in Teradata connector? And if so, can it be parameterized? If memory serves me correctly, you could do this with the old Teradata Enterprise stage that used FastLoad, but we need to use Connector.
Thanks!
Brad