IBM Books

Administration Guide


Appendix P. Splitting Data with db2split

You use the db2split program to split data across database partitions in a partitioned database. The program can be used in two ways:

To process data with db2split, the data must be a form eligible for use by the LOAD utility (except for IXF) data). Date, Time, and Timestamp data must be in the format:

  DATE      YYYY-MM-DD
  TIME      HH.MM.SS
  TIMESTAMP YYYY-MM-DD-HH.MM.SS.XXXXXX

For more information on the LOAD utility, see "Using the LOAD Utility".

The db2split program requires the data file, a configuration file and, optionally, an input partitioning map. (If a customized partitioning map was created, it must be used.) The configuration file contains such information as the name of the input file, the position and length of the partitioning key, and the name of the log file.

The partitioning map file must be generated for the set of nodes for your database (or particular nodegroup, if you created one). A partitioning map file is created when you create a nodegroup, or you can use db2split to create one, or you can use the db2gpmap program to obtain one from the system catalogs and specify that the db2split use it. The db2split program appends each row (or record) to the output file destined for the database partition at which this row should reside, as indicated by the partitioning map.

The db2split program supports codepage conversion. There are two parameters used with the program to specify the source codepage and the target codepage.


[ Top of Page | Previous Page | Next Page | Table of Contents | Index ]

[ DB2 List of Books | Search the DB2 Books ]