Parallel Relational Database Systems Research Articles

Learning about Transitions: Adaptive Control in the Molecular Marshal (M2) Framework

Improvements in sampling on the events leading to transitions can provide significant insights into what drives biomolecular change. We present additions to our Molecular Marshal (M2) software framework for the adaptive sampling of biomolecular transitions. As an example, we work with the transitions seen in BPTI from DEShaw Research (Science, 2010) using their long-running 1 ms trajectory as the basis for states and transitions analyzed from the Anton production trajectory. Our algorithm works by resampling snapshot conformations known to be right before transition events, creating an ensemble set of transitions preconditioned on sampling in the space near to the transitions. This enables us to explore the reduced degrees of freedom that drive the transitions and to examine the statistical foundations of the description for state transitions in BPTI. To achieve these goals we use an adaptive framework for resampling built around a parallel relational database system and with scripts controlling molecular dynamics codes running on XSEDE sponsored national supercomputers. In addition to BPTI, we will show results that start from other trajectories defined from peptides and from other long-running protein simulations. Our scripts for the initial analysis and the resampling thus readily generalize to both long and short trajectory runs and can be used to increase sampling on a broad range of transition events.

Алгоритм репрезентативного сэмплинга для систем баз данных на основе фрагментного параллелизма

Sampling is a popular approach to very large databases processing in a wide range of applications, e.g. data mining, histograms construction, query execution cost estimation, etc. Use of either the sample instead of the original database can reduce the accuracy of the results, but offset by a reduction of time executing processing. Representative sampling allows you to save the sample of certain characteristics of the database. However, existing algorithms for representative sampling can not be used for pas-parallel database systems because it does not take into account the characteristics of the data distribution fissionable by the compute nodes of the cluster system. In this paper we propose al-representative sampling algorithm for parallel relational database systems based on the slice of parallelism. The results of computational experiments on the proposed algorithm, showing adequate maintenance of representativity database properties distributed across the nodes of a cluster system.

Learning About Transitions: Adaptive Controls for the Molecular Dynamics Database

Improvements in sampling on the events leading to transitions can provide significant insights into what drives biomolecular change. We present additions to our Molecular Dynamics Database (MDDB). We start with the transitions seen in BPTI from DEShaw Research (Science, 2010) using their long-running 1 ms trajectory as the basis for a Markov chain with states and transitions analyzed from the Anton production trajectory. Our algorithm works by resampling snapshot conformations known to be right before transition events, creating an ensemble set of transitions preconditioned on sampling in the space near to the transitions. This enables us to explore the reduced degrees of freedom that drive the transitions and to examine the statistical foundations of the Markov chain description for state transitions in BPTI. To achieve these goals we use an adaptive framework for resampling built around a parallel relational database system and with scripts controlling molecular dynamics codes running on XSEDE sponsored national supercomputers. In addition to BPTI, we will show results that start from other trajectories defined from peptides and from other long-running protein simulations. Our scripts for the initial analysis and the resampling thus readily generalize to both long and short trajectory runs and can be used to increase sampling on a broad range of transition events.

Accelerating large semantic web databases by parallel join computations of SPARQL queries

While a number of optimizing techniques have been developed to efficiently process increasing large Semantic Web databases, these optimization approaches have not fully leveraged the powerful computation capability of modern computers. Today's multi-core computers promise an enormous performance boost by providing a parallel computing platform. Although the parallel relational database systems have been well built, parallel query computing in Semantic Web databases have not extensively been studied. In this work, we develop the parallel algorithms for join computations of SPARQL queries. Our performance study shows that the parallel computation of SPARQL queries significantly speeds up querying large Semantic Web databases.

Analytical response time estimation in parallel relational database systems

Techniques for performance estimation in parallel database systems are well established for parameters such as throughput, bottlenecks and resource utilisation. However, response time estimation is a complex activity which is difficult to predict and has attracted research for a number of years. Simulation is one option for predicting response time but this is a costly process. Analytical modelling is a less expensive option but requires approximations and assumptions about the queueing networks built up in real parallel database machines which are often questionable and few of the papers on analytical approaches are backed by results from validation against real machines. This paper describes a new analytical approach for response time estimation that is based on a detailed study of different approaches and assumptions. The approach has been validated against two commercial parallel DBMSs running on actual parallel machines and is shown to produce acceptable accuracy.

The impact of load balancing to object-oriented query execution scheduling in parallel machine environment

Skewness has been one of the major problems not only in parallel relational database systems, but also in parallel object-oriented database systems. To improve performance of object-oriented query processing, a careful and intelligent skew handling for load balancing must be established. Depending on the parallel machine environment, whether it is a shared-memory or a shared-nothing architecture, load balancing can be achieved through “physical” or “logical” data re-distribution. It is not the aim of this paper to propose or to investigate skew handling methods, but rather to analyze the impact of load balancing to query execution scheduling strategies. Our analysis shows that when load balancing is achieved, “serial” execution scheduling is preferable to “parallel” execution scheduling strategy. In other words, allocating full resources to a sub-query seems to be better than dividing resources to multiple sub-queries.

Data placement in a parallel DBMS with multiple disks

Various strategies have been developed to assist in determining an effective data placement for a parallel database system. However, little work has been done on assessing the relative performance obtained from different strategies. This paper studies the effects of different data placement strategies on a shared-nothing parallel relational database system when the number of disks attached to each node is varied. Three representative strategies have been used for the study and the performance of the resulting configuration has been assessed in the context of the TPC–C benchmark running on a specific parallel system. Results show an increase in sensitivity to data placement strategy with increasing number of disks per node.

Uniform partitioning of relations using histogram equalization framework: An efficient parallel hash-based join

Many parallel join algorithms have been proposed for parallel relational database systems. Among them, the parallel hash-based join algorithm has been found to be superior to other ones for the uniform distribution of data. This paper proposes an efficient algorithm, called skew resolution join algorithm (SRJA), for parallel join algorithms with skewed data. We propose a methodology for partitioning relations evenly across all processors in a parallel database system. We performed simulation experiments under the Zipf-like distribution of hashed values for join attributes.