Gregory D. Speegle

Formal aspects of concurrency control in long-duration transaction systems using the NT/PV model

In the typical database system, an execution is correct if it is equivalent to some serial execution. This criterion, called serializability, is unacceptable for new database applications which require long-duration transactions. We present a new transaction model which allows correctness criteria more suitable for these applications. This model combines three enhancements to the standard model: nested transactions, explicit predicates, and multiple versions. These features yield the name of the new model, nested transactions with predicates and versions, or NT/PV. The modular nature of the NT/PV model allows a straightforward representation of simple systems. It also provides a formal framework for describing complex interactions. The most complex interactions the model allows can be captured by a protocol which exploits all of the semantics available to the NT/PV model. An example of these interactions is shown in a CASE application. The example shows how a system based on the NT/PV model is superior to both standard database techniques and unrestricted systems in both correctness and performance.

Introduction to JDBC

JDBC is an API defined in the java.sql and javax.sql packages for connecting an arbitrary database to a Java program. This chapter uses an analogy to understand the basic working of JDBC. It supposes that one operates a small business that sells gadgets. The gadgets are produced in a factory, across a river from the store. Without a boat or a bridge, there is no way to get to the factory or for the factory to deliver goods to the store. This represents the situation with a database and a Java program: the Java program is the store, and the database is the factory. Without something to help, there is no way for the Java program to make requests of the database, and the database cannot communicate with the Java program. It would be logical to build a bridge over the river to transport the products from the factory to the store. The analogous piece of software for the Java program and database is an interface called the driver. The driver is loaded by setting the jdbc.drivers property of the Java virtual machine (JVM). As modern rivers are crossed by many bridges, there are many drivers to connect Java programs to databases.

Long-duration transactions in software design projects

An example of a software development application is considered, and the formal model of H. Korth and G. Speegle (1988) is applied to show how this example could be represented as a set of database transactions. It is shown that, although the standard notion of correctness (serializability) is too strict, the notion of correctness in the Korth and Speegle model allows sufficient concurrency with acceptable overhead. An extrapolation is made from this example to draw some conclusions regarding the potential usefulness of a formal approach to the management of long-duration design transactions.<<ETX>>

A Meta-Structure for Supporting Multimedia Editing in Object-Oriented Databases

Multimedia databases include many types of new data. One common property of these data items is that they are very large. We exploit the concept of transformational representations to logically store images without the bitmap. This technique is similax to views in traditional database systems. The technique is useful when users axe editing images in the database to create new images. Our method produces significant space savings over already compressed images, and potentially greater savings for video and audio. The method requires the support of multimedia editors. This paper emphasizes the meta-structure needed by the database to support multimedia editing.

Views of media objects in multimedia databases

Multimedia database systems must manage large amounts of graphical, textual, and audio data. To compound the problem, if a multimedia designer creates a new media image from an existing item, current systems store the new object as an independent image. This space is wasted since the data already exists in the system. In order to eliminate this wasted space, views of stored objects should be used instead of replicated data. Views are well known in relational databases to have both benefits and drawbacks. In order to overcome these drawbacks, the paper defines a view schema, which mirrors the media hierarchy; a view dag which determines the base files used to create the view; and view independence, a paradigm for performing operations on views. The criteria for a canonical media algebra are outlined. A canonical media algebra allows multiple editors to instantiate a view. Finally, optimization of specifications is discussed.

Extending databases to support image editing

In order to understand similarity between images, recent research has focused on adaptable searches (Seidl and Kreigel, 1997) and fuzzy queries (Fagin, 1998). However, one of the best means for determining similarity between images is to know how the image was created (Brown et al., 1999). If the image is a combination of other images in the database, then there is a great deal of similarity between the base images and the created one. This requires extending the database to support image editing operations. We have built a preliminary system which does this by using a Web-based image editor and an image server. The editor and the server understand a logical model language that represents images. This paper then explores the issues of performance for deriving images from a sequence of operations.

Issues in using specifications to improve content-based search of multimedia data

Many current multimedia database management systems perform content-based retrieval of images by extracting the values of various features from every object stored in their system. This can be time-consuming, especially if extracting the features requires a human to analyze each object. This process can be minimized in multimedia database systems that store images as a sequence of editing operations, called specifications, instead of the usual binary format. This paper discusses the advantages and applicability of such systems and the issues that must be resolved in order to develop them.

Quantifying the benefits of semantics

An active area of current research is the use of semantics in concurrency control. Simulations using a new concurrency control protocol, called complex two-phase locking, can quantify the benefits of using semantics within long-duration transaction systems. It is then possible to determine if the benefits gained are worth the human effort required to obtain the semantics.

Creating, Deleting, and Altering Tables

Tables and all other parts of a database that are not data are collectively known as metadata. Metadata is data about data. Structured query language (SQL) is also known as metadata. The basic table has a name and a set of columns, each with its own data type. One can create a table in SQL using CREATE TABLE. This creates a table named . The columns of the table are specified in a comma-delimited list of name/data type pairs. Before creating a table, most database management systems (DBMSs) require the creation of a database to hold the new table. When a new row is created, any columns without a specified value are assigned the default value. Unless otherwise specified, the default value of a column is NULL. SQL allows specifying a default value for a column using the DEFAULT clause. DBMS can do much more than just store and access data. It can also enforce rules (called constraints) on what data are allowed in the database. Such constraints are important because they help maintain data integrity. SQL enforces constraints by prohibiting any data in the database that violate any constraint. Any insert, update, or delete that would result in a constraint violation is rejected without changing the database.

Introduction to Cursors, Embedded SQL, Stored Procedures, and Triggers

SQL is powerful, but it also requires more capabilities to make it truly useful. Such capabilities include executing SQL and accessing results in other programming languages, scripting procedures, and reacting to changes in the database. An SQL query returns an entire set of rows. In some instances, one may wish to process a result one row at a time instead of all at once. This can be done through a cursor. A cursor is basically a pointer to some position within the rows of a result set. Sometimes single SQL statements are not sufficient for what one needs to do. There are several ways to create a sequence of SQL statements to solve a problem, such as stored procedures. SQL allows the creation of scripts, called stored procedures, within the database. These scripts may contain one or more SQL statements. Stored procedures may take parameter values as input and even return results. The scripting language also includes loops, conditionals, and variables.