Browse Prior Art Database

Alternative Method of Implementing Long Name for a Large Database Management System

IP.com Disclosure Number: IPCOM000021711D
Original Publication Date: 2004-Feb-04
Included in the Prior Art Database: 2004-Feb-04
Document File: 2 page(s) / 31K

Publishing Venue

IBM

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 2

Alternative Method of Implementing Long Name for a Large Database Management System

Disclosed is an efficient, economical method to implement a long name in a software product.

Traditionally, the lengths of names used in a large database management system(DBMS) are limited to 8 maximum for short identifiers and 18 maximum for long identifiers. With the prevalent use of personal computers, mid-range computers to communicate with a mainframe system, and other factors, the need arises for names to expand to longer size. A large database management system has to handle the task of processing names as long as 128 bytes or longer.

The new Distributed Relational Database Architecture (DRDA) allows names up to 255 characters long because large database management systems such as DB2 UDB for z/OS or Oracle are complex and comprised of many millions lines of code that process the names or do processing based on the names. Expanding the names to a larger length and maintain functionality is a daunting task if not designed correctly. If not designed cleverly, a large amount of resources will be used to expand the fields in each component in the DBMS and many lines of code will have to be written to accommodate the longer length. Ninety percent of customers can get by without using the long name(names greater than the traditional length), so the return of investment is not proportional to the resource used and does not make good business sense.

To enable a large DBMS to accept long names as identifiers, the traditional method being used by certain large DBMS is to mechanically expand all the name fields in all the components to the new maximum of 128 characters. Because the name fields are many and the program logic that use them are abundant, numerous places in the modulesneed change. Consequently, this mechanical method consumes a lot of resources and it is time-consuming.

The method proposed uses a pseudonym for long names. As soon as the long names enter the DBMS, they are translated into pseudonyms and stored in a long-name-psuedonym table. This table is accessible to all components that have a need to use them - such as when generating an error message or checking the DB2 catalog for authority, or in a distributed environment when the data is ready to be sent across the wire to a remote site. When these components have a need to use the long name, they use the pseudonym to get the long name out. This pseudonym does not even have to be a real name and its length...