A logical-level approach to data base conversion

  • Authors:
  • Arie Shoshani

  • Affiliations:
  • System Development Corporation, Santa Monica, California

  • Venue:
  • SIGMOD '75 Proceedings of the 1975 ACM SIGMOD international conference on Management of data
  • Year:
  • 1975

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we describe an ongoing project which is addressing the problem of converting and transferring data bases among disparate data management systems (DMSs). The difficulties in converting a data base from one DMS to another stem from the fact that data base structures are system and application dependent. As a result, data base structures embed constraints of three types: (1) logical-level constraints, such as hierarchies, networks, size and type of fields; (2) storage-level constraints, such as inversion capabilities, access paths, and indexing organization; and (3) physical-level constraints, such as physical devices and block/record structures.The approach taken by this project is based on the concept that the data conversion process can depend basically on conversion at the logical level only. Conversion at this level can be achieved by using existing query and generate capabilities of DMSs to move data from their physical representation to the logical level and vice versa. Detailed descriptions of the system components and the languages supporting them are given.An important aspect of the work is in the area of semantics of logical data conversion. We choose to restrict the basic conversion types to field-to-field mappings. Then, semantic analysis determines which combinations of conversion types are permissible. This approach allows for conversion tools that are powerful yet simple to specify.Finally, some observations on the implications of data conversion needs for the design of data management systems are suggested.