Annita N. Wilschut

Dataflow query execution in a parallel main-memory environment

In this paper, the performance and characteristics of the execution of various join-trees on a parallel DBMS are studied. The results of this study are a step into the direction of the design of a query optimization strategy that is fit for parallel execution of complex queries.Among others, synchronization issues are identified to limit the performance gain from parallelism. A new hash-join algorithm is introduced that has fewer synchronization constraints than the known hash-join algorithms. Also, the behavior of individual join operations in a join-tree is studied in a simulation experiment. The results show that the introduced Pipelining hash-join algorithm yields a better performance for multi-join queries. The format of the optimal join-tree appears to depend on the size of the operands of the join: A multi-join between small operands performs best with a bushy schedule; larger operands are better off with a linear schedule. The results from the simulation study are confirmed with an analytic model for dataflow query execution.

PRISMA/DB: a parallel, main memory relational DBMS

PRISMA/DB, a full-fledged parallel, main memory relational database management system (DBMS) is described. PRISMA/DBs high performance is obtained by the use of parallelism for query processing and main memory storage of the entire database. A flexible architecture for experimenting with functionality and performance is obtained using a modular implementation of the system in an object-oriented programming language. The design and implementation of PRISMA/DB are described in detail. A performance evaluation of the system shows that the system is comparable to other state-of-the-art database machines. The prototype implementation of the system runs on a 100-node parallel multiprocessor. >

Parallel evaluation of multi-join queries

A number of execution strategies for parallel evaluation of multi-join queries have been proposed in the literature; their performance was evaluated by simulation. In this paper we give a comparative performance evaluation of four execution strategies by implementing all of them on the same parallel database system, PRISMA/DB. Experiments have been done up to 80 processors. The basic strategy is to first determine an execution schedule with minimum total cost and then parallelize this schedule with one of the four execution strategies. These strategies, coming from the literature, are named: Sequential Parallel, Synchronous Execution, Segmented Right-Deep, and Full Parallel. Based on the experiments clear guidelines are given when to use which strategy.

Pipelining in query execution

It is shown that the use of an adjusted algorithm can improve the performance of a nested query execution containing join operations. Apart from the equijoin, the union, Cartesian product, Theta -join, and intersection operations can be implemented in a pipelining way. The difference, however, cannot be implemented in this way. The proposed algorithms are implemented in PRISMA/DB.<<ETX>>

Road collapse in Magnum

This paper describes the implementation of a triangulation based collapse algorithm in the general-purpose object oriented DBMS Magnum. The contribution of the paper is twofold. First, we show that true integration of complex spatial functionality in a DBMS can be achieved. Second, we worked out a collapse algorithm to be used in the complex area of map generalization.

Parallel query execution in PRISMA/DB

1 I n t r o d u c t i o n In the PRISMA-project, a large multi-processor system has been built, is be used to study the performance gains from parallelism. A parallel, main-memory relational database system (PItlSMA/DB) runs on this so-called POOMA-machine. This paper studies the possibilities of using parallelism to improve the performance of relational database management systems. Because the equi-join is an important, and time-consuming operation, queries consisting of a number of equi-joins are used to describe how different forms of parallelism can speed up the execution of such queries. This paper is organized as follows: First, a brief introduction into PRISMA and the DBMS running on it is given. After that, different forms of parallelism are described and the ways in which they can be used is identified. Using this knowledge, the possible parallelism in the execution of join-queries is discussed. Special attention is paid to pipelining. It is shown, that pipelining needs a new hash-join algorithm and that using this algorithm may yield effective parallelism over a pipeline of join operations. Finally, we discuss the implications of using pipelining as a source of parallelism for the optimization of join queries. The paper is concluded with our plans for future work. The PaRallel Inference and Storage MAchine PRISMA is a highly parallel machine for data and knowledge processing. The PRISMA-ma~hine contains 100 nodes that each contain a data processor, a communication processor and 16 Mbyte of local memory. 50 nodes have a disk a~d some nodes have an ethernet card that provides an interface with a host computer. Each communication processor connects a node to 4 other nodes. In this way a fast, high-bandwidth network is provided. This hardware can be classified as a shared-nothing multi-processor system. The maz~ine is designed to support a relational main memory database management system PRISMA/DB. An extensive introduction to this system can be found in [Kers87] and in [Wils89]. Here, only the features that are important for this paper are summarized. PRISMA/DB stores the entire database in main memory. Disks are used for backup only. To gain performance and to make storage in main memory feasible, the tuples belonging to one relation are fragmented over more than one node. A fragment is a set of tuples that belong to the same relation and that reside on the same node. A relation does not necessarily use all available nodes. …

Database research at the University of Twente

The University of Twente is one of the leading technical universities in the Netherlands. It is the only Dutch university that has an independent computerscience department. The database research group is comprised of approximately 15 researchers and graduate students. It is participating in a number of externally funded, national and international research projecs, and has done so in the recent past. The research interests of the database group are in the field of object-oriented, logical and extended relational systems. The three main points of interest are: i) development of well-founded (objed-oriented) data models and tools, ii) design of and optimization in database systems, with a strong emphasis on distribution aspects, and iii) database support for new (technical) applications. In the following, We will give a brief description of our main research efforts, and mention some of the associated publications.

Parallel Evaluation of Multi-join Queries

A number of execution strategies for parallel evaluation of multi-join queries have been proposed in the literature. In this paper we give a comparative performance evaluation of four execution strategies by implementing all of them on the same parallel database system, PRISMA/DB. Experiments have been done up to 80 processors. These strategies, coming from the literature, are named: Sequential Parallel, Synchronous Execution, Segmented Right-Deep, and Full Parallel. Based on the experiments clear guidelines are given when to use which strategy.