this paper, we discuss the principles, architecture, and implementation of the Penguin approach to sharing persistent objects. The primary motivation for using a database management system (DBMS) is to allow sharing of data among multiple customers and multiple applications. To support sharing among independent transactions, DBMSs have evolved services including transaction independence, persistence, and concurrency control. When a database is shared among multiple applications, these applications typically have differing requirements for data access and representation. Such differences are supported by having views, which present diverse subsets of the base data [ChamGT75]. The primary motivation for defining objects is to include sharable semantics and structure in the information. Must all applications sharing objects use the same object schema, or is it better to give each application its own object schema and somehow to integrate them? If multiple applications differ in view the needed compromise reduces the relevance and effectiveness of the object representation [AbitB91]. For instance, customers will have an orthogonal view of an inventory versus the suppliers. 2 When combining independently developed applications, we do not have the luxury of choosing a common object schema. Many legacy databases and legacy data are still being used. We must retain the investment in existing application software and databases, while building new software using the object approach. When creating a federation of heterogeneous (pre-existing) databases, we must support a degree of interoperation among these databases and their schemas. Consider also that current projects will become legacy in a few years hence, but their semantics will remain. Whatever solutions we create in shar...

We describe a two-level client-side cache for composite objects mapped as views of a relational database. A semantic model, the Structural Model, is used to specify joins on the relational database that are useful for defining composite objects. The lower level of the cache contains the tuples from each relation that have already been loaded into memory. These tuples are linked together from relation to relation according to the joins of the structural model. This level of the cache is shared among all applications using the data on this client. The higher level of the cache contains composed objects of data extracted from the lower level cache. This level of the cache uses the object schema of a single application, and the data is copied from the lower level cache for convenient access by the application. This two-level cache is designed as part of the Penguin system, which supports multiple applications, each with its own object schema, to share data stored in a common relational dat...

This dissertation discusses the realization of object persistence, in a CORBA-based distributed system, through the integration of the Object Request Broker (ORB) with an Object Database Management Systems (ODBMS). Three approaches to persistence of CORBA objects are described. The first one, pseudopersistence, makes object references persistent, but not transparently storable. The others, smart pointerbased persistence and virtual persistence, provide transparent storability to object references, both in the case of references to local objects and in the case of references to remote objects. We stress the ORB and ODBMS features each approach depends upon, and point out the ones not fully standardized at the present time.

Due to the wide use of object-oriented technology in software development, reverse engineering of relational schemas to object-oriented schemas is gaining a lot of interest. One of the major problems with existing approaches for this schema mapping is the extensive amount of information that must be gathered either automatically or from the user. This paper presents an object-centered approach (as opposed to other relation-centered approaches) for doing the schema mapping. The procedure maps a 3NF relational schema into an object-oriented schema without the explicit use of inclusion dependencies and provides a greater scope for automation. Introduction Object-oriented technology is being widely used in software development these days. The object-oriented data model is growing in popularity in the area of database development (Catell 1991). Consequently, there is also a growing interest in re-engineering relational databases to object-oriented databases. Reverse engineering of relationa...

We present the query processing algorithm for the Penguin system. Penguin supports multiple object views on a relational database, so that data may be shared by applications with heterogeneous object schemata. Penguin also offers an interface for the C++ language. We have developed a query processing algorithm for querying composite object views. The query algorithm takes a query on a composite object view, and decomposes it into partial queries on the relational database. The final query result is composed from those partial results. We also discuss an optimization for reducing the volume of temporary data created in query processing. 1 Introduction Object-oriented programming is becoming quite prevalent for its benefits of software productivity, quality and reusability. Software developers are eager to adopt the object-oriented approach in new applications. On the other hand, the relational model introduced by Codd [7] is so widely accepted that most new database applications use re...

this document is organized as follows. A literature review covering the area of reverse engineering of relational schemas to other conceptual schemas is given in Chapter II. Chapter III establishes the context for this dissertation research. It includes the definitions of frequently-used terms, a discussion on the 9 various alternatives that were considered to address the research problem, and an overview of the entire approach. Chapter IV describes in detail the schema mapping and data mapping components of SOAR. The chapter includes all the main algorithms required to carry out the schema mapping process. A description of a data mapping procedure is also presented in the chapter. The evaluation of the mapping algorithms based on consistency and completeness is given in Chapter V. In order to illustrate the capabilities of the mapping procedures described in this dissertation, the results of mapping four relational schemas are given in Chapter VI. Finally, the conclusions and potential for future work are discussed Chapter VII. The appendices include the file format of the mapping database ("meta-database") and a sampling of the source files generated by SOAR. CHAPTER II LITERATURE REVIEW This chapter presents a comprehensive literature review in areas related to research in providing object-oriented access to existing relational databases. The chapter is divided into three sections. The first section looks at work in the area of schema mapping from the relational model to another semantic model such as the entity-relationship (ER) model, the extended ER (EER) model, or the objectoriented model. The next section looks at some research projects whose objective was the development of an integrated approach. The final section contains references to work related to thi...

We present the implementation of the query function for the Penguin system. Penguin is an object-oriented database system that supports multiple object views on a relational database. It enables many applications to share a database using different object schemata. Also, users can take queries for the Penguin database in their applications, to retrieve objects on the heterogeneous data model. The query function is very powerful for manufacturing and engineering purposes, that is, connections between relations and methods defined by users in applications are available in query requests. We introduce the object model and the query execution algorithm of the Penguin system with an example for a steel plant, and show the system configuration for the implementation of the query function. 1 Introduction In the field of manufacturing and engineering, an important requirement for database systems is the handling of complicated structures of data, such as manufacturing design data and producti...