问题
So I know that Teradata has a limitation of 64k bytes per row. I have a wide table that I need to export to Teradata and there are some fields (varchar (5000)) that go along with that. We have seen cases where the row size exceeds this limitation.
So, my question is, how can we overcome this situation? We cannot trim the large VARCHARS in our source as they are necessary to the downstream business users. Splitting up the table is always the option, but are there any other ways in Teradata that this can be handled?
回答1:
Teradata 16.00 increases row size to 1MB. Otherwise, as JNevill mentions split the table into two with the same primary index. This allows them to be joined together AMP-local on a row-key merge join, one of the most efficient joins in Teradata.
来源：https://stackoverflow.com/questions/52228089/row-size-limitation-in-teradata