16.20 - Type 3: Batch - Teradata SQL Assistant

Teradata® SQL Assistant for Windows User Guide

Product
Teradata SQL Assistant
Release Number
16.20
Published
October 2018
Language
English (United States)
Last Update
2020-01-03
dita:mapPath
gfk1537201040714.ditamap
dita:ditavalPath
ft:empty
This form applies when the following conditions are met:
  • Teradata SQL Assistant is connected through Teradata.NET or Oracle.net, not ODBC.
  • The query consists of a single INSERT, UPDATE, or DELETE statement.
  • The statement must not contain any LOB parameters [?B or ?C].
  • The Batch size in the Import tab of the Options dialog box is not specified as one.

If the Batch size is set to one, the Batch import mode is not used. For larger values, the data provider packs the data for the specified number of rows into a batch (as long as it fits into a 64k parcel), and executes one batch at a time.

In general, the larger the batch size, the better the performance. With larger batch sizes:
  • Data for a single batch must be less than 1MB.
  • The entire Import file is loaded into memory.
  • If a batch import results in an error, the entire batch is rolled back and Teradata SQL Assistant reports which record caused the error.
  • If multiple records result in errors in a single batch, the indicated record might not be the first problem record in the file.

The Silently discard rows that cause constraint violations option on the Import tab of the Options dialog can be selected if the data is just a sample, or if loss of duplicates or other rows that cause constraint violations are not a problem. If this option is selected, all errors are ignored and Teradata SQL Assistant reports the number of rows that were imported and the number of rows discarded.