Fragkiskos Pentaris

Query optimization in distributed networks of autonomous database systems

Large-scale distributed environments, where each node is completely autonomous and offers services to its peers through external communication, pose significant challenges to query processing and optimization. Autonomy is the main source of the problem, as it results in lack of knowledge about any particular node with respect to the information it can produce and its characteristics, for example, cost of production or quality of produced results. In this article, inspired by e-commerce technology, we recognize queries as commodities and model query optimization as a trading negotiation process. Subquery answers and subquery operator execution jobs are traded between nodes until deals are struck with some nodes for all of them. Such trading may also occur recursively, in the sense that some nodes may play the role of intermediaries between other nodes (subcontracting). We identify the key parameters of the overall framework and suggest several potential alternatives for each one. In comparison to trading negotiations for e-commerce, query optimization faces unique new challenges that stem primarily from the fact that queries have a complex structure and can be broken into smaller parts. We address these challenges through a particular instantiation of our framework focusing primarily on the optimization algorithms run on “buying” and “selling” nodes, the evaluation metrics of the queries, and the negotiation strategy. Finally, we present the results of several experiments that demonstrate the performance characteristics of our approach compared to those of traditional query optimization.

A Statistical Metadata Model for Simultaneous Manipulation of both Data and Metadata

There is a growing demand for more cost-efficient production processes in Statistical Institutes. One way to address this need is to equip Statistical Information Systems (SIS) with the ability to automatically produce statistical data and metadata of high quality and deliver them to the user via the Internet. Current approaches, although provide for the storage of appropriate metadata, do not use process metadata for guiding the production process. In this paper we present an approach on creating SISs that permits metadata-guided statistical processing based on an object-based, statistical metadata model. The model is not domain specific and can accommodate both microdata and macrodata. We have equipped the model with a set of transformations that can be used to automatically manipulate data and metadata. We show the applicability of transformations with some examples using actual statistical data for R&D expenditures. Finally, we demonstrate how the presented framework can be exploited for the construction of a web site that offers ad hoc query capabilities to the users of statistical data.

Autonomic Query Allocation based on Microeconomics Principles

In large federations of autonomous database systems, automatic distribution of the query workload to those systems is a critical issue. We examine this problem under the perspective of microeconomics theory and show how the latter can be used to construct an efficient decentralized mechanism that maximizes system throughput. In particular, we introduce a solution that is based on the notion of query markets. We examine the properties of these markets and show that they result in Pareto-optimal allocations of resources to queries. An extensive set of experiments with both a simulator and an actual implementation on top of a commercial DBMS demonstrate significant improvements in the overall system throughput when our technique is used.

A statistical metadata model for clinical trials' data management

We introduce a statistical, process-oriented metadata model to describe the process of medical research data collection, management, results analysis and dissemination. Our approach explicitly provides a structure for pieces of information used in Clinical Study Data Management Systems, enabling a more active role for any associated metadata. Using the object-oriented paradigm, we describe the classes of our model that participate during the design of a clinical trial and the subsequent collection and management of the relevant data. The advantage of our approach is that we focus on presenting the structural inter-relation of these classes when used during datasets manipulation by proposing certain transformations that model the simultaneous processing of both data and metadata. Our solution reduces the possibility of human errors and allows for the tracking of all changes made during datasets lifecycle. The explicit modeling of processing steps improves data quality and assists in the problem of handling data collected in different clinical trials. The case study illustrates the applicability of the proposed framework demonstrating conceptually the simultaneous handling of datasets collected during two randomized clinical studies. Finally, we provide the main considerations for implementing the proposed framework into a modern Metadata-enabled Information System.

Distributed Query Optimization by Query Trading

Large-scale distributed environments, where each node is completely autonomous and offers services to its peers through external communication, pose significant challenges to query processing and optimization. Autonomy is the main source of the problem, as it results in lack of knowledge about any particular node with respect to the information it can produce and its characteristics. Internode competition is another source of the problem, as it results in potentially inconsistent behavior of the nodes at different times. In this paper, inspired by e-commerce technology, we recognize queries (and query answers) as commodities and model query optimization as a trading negotiation process. Query parts (and their answers) are traded between nodes until deals are struck with some nodes for all of them. We identify the key parameters of this framework and suggest several potential alternatives for each one. Finally, we conclude with some experiments that demonstrate the scalability and performance characteristics of our approach compared to those of traditional query optimization.

Mapping objects

We present a technique that is based on volatile mapping objects and enables wrappers-based mediation architectures to describe bi-directional (read–write) interschema mappings of multiple, disparate data sources. We describe the structure of these mapping objects, explain how they work, and compare them to other traditional techniques used for describing schema mappings in data-mediation systems.

Modeling statistical metadata

An object oriented statistical metadata model is presented, which can be used in building information systems providing metadata-guided, statistical data processing features. The semantics of the model are analyzed and a set of operators (transformations) is proposed that allows for the automatic manipulation of both data and metadata at the same time. We discuss the mathematical properties of these transformations, and subsequently as a case study, we demonstrate how a statistical office can use the presented framework to build a Web site offering ad hoc query capabilities to its data consumers.

Self-deadlocks in disparate scientific data management systems

In large statistical and scientific data management environments, where mediation architectures are used to integrate disparate and autonomous systems, a new problem - self-deadlock - may cause global transaction failures. In this short paper we briefly examine the reasons causing this problem and identify some algorithms for resolving it.

Query trading in digital libraries

In this paper we present a distributed query framework suitable for use in federations of digital libraries (DL). Inspired by e-commerce technology, we recognize CPU-processing and queries (and query answers) as commodities and model the task of query optimization and execution as a task of trading CPU-processing, queries and query-answers. We show that our framework satisfies the needs of modern DL federations by respecting the autonomy of DL nodes and natively supporting their business model. Our query processing conception is independent of the possible distributed architecture and can be easily implemented over a typical GRID architectural infrastructure or a Peer-To-Peer network.