Although many extended transaction models have been proposed [Elm93], few practical implementations exist and even fewer can support more than one model. We present the Reflective Transaction Framework, as a practical and modular method to implement extended transaction models. We achieve modularity by applying the Open Implementation approach [Kic92] (also known as meta-object protocol [KdRB91]) to the design of the reflective transaction framework. We achieve practicality by implementing on top of a commercial transaction processing monitor. For our implementation of the reflective transaction framework, we introduce transaction adapters, add-on modules built on top of existing commercial TP components, such as Encina, that extend their functionality to support extended transaction features and semantics. Since our framework design is based on the transaction processing monitor architecture [GR93], it is widely applicable to many modern TP monitors. The reflective transaction framew...

We map an object model to a commercial relational multi-processor database system using replication and view materialisation to provide fast retrieval. To speed up complex update operations, we exploit intra-transaction parallelism by breaking such an operation down into shorter relational operations which are executed as parallel subtransactions of the update transaction. To ensure the correctness and recoverability of the operation 's execution, we use multi-level transactions. In addition, we minimise the resulting overhead for the logging of the compensating inverse operation required by the multi-level concept by logging the compensation for non-derived data only. In particular, we concentrate on the novel application of multi-level transaction management to efficiently maintain the replicated data and materialised views. We present a prototype implementation and give a performance evaluation of an exemplary set-oriented update statement. 1 Introduction During the past decade, r...

This paper reports on results and experiences from the COMFORT automatic tuning project. The objective of the project has been to investigate architectural principles of self-tuning database and transaction processing systems, and to develop self-tuning methods for specific performance tuning problems. A particular concern of the project has been to cope with workload dynamics and workload heterogeneity in multi-user systems. As a general guideline, an adaptive feedback control approach has been adopted, where observations of the current load characteristics are used to predict performance trends and to drive the dynamic adjustment of tuning parameters. As examples of these general principles, the paper discusses adaptive approaches to two specific tuning problems and the developed solutions. First, we present a self-tuning load control method that copes with overload caused by excessive lock conflicts that may occur during load surges. This conflict-driven load control method adapts t...

In recent years, databases have started to be used as intelligent repositories for a variety of semantically-richer systems. A consequence of such architectures is that transaction scheduling takes place throughout composite systems consisting of layered subsystems. Such transaction architectures have been studied extensively. Existing theory, however, limits the degree of parallelism, and makes a number of simplifying assumptions which cannot be taken for granted in practice. This paper proposes a new model and correctness criterion, stack conflict consistency, for composite transactional systems. The main contribution of the new model is to establish the correctness conditions under which higher degrees of parallelism can be achieved between operations of the same transaction, as well as between conflicting operations of different transactions, in a uniform way. This possibility, although hinted at previously, has not yet been exploited in practical composite systems. Hence, we hope...

ion Layer Object Manager Exported Database Functionality Win-NT Navigator Distribution Layer Functional Layer Execution Process Development Manager Canonical Representation External INTERFACE SERVICES Development Support PROCESS SERVICES Dispatcher Desktop Interface Fig. 3.1. System architecture of OPERA This canonical representation is not suitable, however, for either commercial databases or user interaction. The database abstraction layer translates the canonical representation to the private representations of the underlying repositories (SQL, C++, system calls) as required by the physical implementation of the underlying database (Figure 3.2). The process service layer contains all the components required for coordinating and monitoring the execution of processes. The most relevant components for the purposes of this paper are the dispatcher, the navigator, the object manager, the query manager, and the exported database functionality module. The dispatcher and the development s...

An iterator is proposed that allows to apply sequences of update operations in a set-oriented way with deterministic semantics. Because the mechanism is independent of a particular model, it can be used in the relational and in object-oriented ones. Thus, the deterministic semantics of embedded SQL cursors, and of triggers that are applied after (set-oriented) SQL updates can be checked. Furthermore, the iterator can be used to apply object-oriented methods, which are usually update sequences defined on a single object, also to sets in a deterministic way. It turns out that the criteria that guarantee determinism are also used in semantic or multi-level concurrency control. 1 Introduction The paper deals with the general problem of defining update languages that are comparable in expressive power to typical query languages. When designing such an update language, one should pursue the following objectives: ffl genericity: update operations should be applicable to all types of objects...

Large knowledge bases that are intended for applications such as CAD, corporate repositories or process control will have to be shared by multiple users. For these systems to scale up, to give acceptable performance and to exhibit consistent behavior, it is mandatory to synchronize user transactions using a concurrency control algorithm. The transactions in knowledge bases often access a large number of entities and perform complex inferences that may last for a long period of time. In such a situation, using conventional concurrency control methods, which require a transaction to hold its locks until it has acquired all the locks it will ever need, do not lead to good performance. This thesis examines the problem of concurrency control for such long transactions in a knowledge base setting. Using a directed graph as a general model of a knowledge base, we develop an algorithm, called the Dynamic Directed Graph (DDG) policy, that allows release of locks by a transaction before it has ...

Large knowledge bases that are intended for applications such as CAD, corporate repositories or process control will have to be shared by multiple users. For these systems to scale up, to give acceptable performance and to exhibit consistent behavior, it is mandatory to synchronize user transactions using a concurrency control algorithm. In this paper, we examine a novel concurrency control policy called Dynamic Directed Graph (or DDG) policy that effectively exploits the rich semantic structure of a knowledge base. Our analysis is carried out in the context of a real knowledge based system application from which knowledge base structure and workload parameters are computed. These serve as a basis for studying the implementation alternatives that arise as a result of knowledge base characteristics. The implementation alternatives that we consider include selection of portions of the knowledge base structure to be exploited for concurrency control, and also the dependence of concurrenc...

We map an object model to a commercial relational multi-processor database system using replication and view materialisation to provide fast retrieval. To speed up complex update operations, we exploit intra-transaction parallelism by breaking such an operation down into shorter relational operations which are executed as parallel subtransactions of the update transaction. To ensure the correctness and recoverability of the operation 's execution, we use multi-level transactions. In addition, we minimise the resulting overhead for the logging of the compensating inverse operation required by the multi-level concept by logging the compensation for non-derived data only. In particular, we concentrate on the novel application of multi-level transaction management to efficiently maintain the replicated data and materialised views. We present a prototype implementation and give a performance evaluation of an exemplary set-oriented update statement. 1 Introduction During the past decade, r...

This paper presents an approach to improving database performance by combining parallelism of multiple independent transactions and parallelism of multiple subtransactions within a transaction. An experimental prototype has been built that supports this combination of inter- and intra-transaction parallelism based on the framework of multi-level transaction management. A performance study for a foreign exchange banking application has been carried out to obtain insights into the interaction of inter- and intra-transaction parallelism. Significant performance improvements could be gained when both update and retrieval transactions were parallelized. 1 Introduction Database transaction processing differs from other parallel applications in their multi-user characteristics. In online transaction processing [GR93], also known as OLTP, many transactions are invoked from a large number of terminals, where each transaction includes a relatively short sequence of database accesses. Usually,...