Data Load

The data-load process reads a source data file, converts the data to a different format, and inserts the converted data into a database table. The source data can come from one or more of the following sources:

• Files
• Tapes
• Pipes (application-generated data)

The HPL supports load and unload data greater than 2 gigabytes to files and tapes.

During conversion, the source data is often manipulated so that the converted data displays different characteristics. Examples of this manipulation include:

• Changing lowercase letters to uppercase letters
• Loading default values, loading certain table columns, or replacing nulls
• Masking the data to include only part of a value
• Converting from one data type to another, such as conversion of a numeric string to a float
• Converting from the code set of one locale to the code set of another locale

When you prepare to run a data load using the HPL, you describe the actions that the HPL must take by defining a set of metadata components. The components describe different aspects of the load process. Figure 1 shows the data load process. The HPL uses information from:

• The device array to find the set of the source-data files
• The format to define the data types and layout of the source data
• The filter to select the records from the source data that should be written to the database table
• The map to modify and reorganize the data

The ipload utility helps you prepare the components. Loading Data to a Database Table, addresses the process of loading a file to a database.

The figure below summarizes the data-load process by showing how data moves from data files to table entries.

Figure 1. The Data-Load Process