Michael Jurczyk

A mathematical model, heuristic, and simulation study for a basic data staging problem in a heterogeneous networking environment

Data staging is an important data management problem for a distributed heterogeneous networking environment, where each data storage location and intermediate node may have specific data available, storage limitations, and communication links. Sites in the network request data items and each item is associated with a specific deadline and priority. It is assumed that not all requests can be satisfied by their deadline. The work concentrates on solving a basic version of the data staging problem in which all parameter values for the communication system and the data request information represent the best known information collected so far and stay fixed throughout the scheduling process. A mathematical model for the basic data staging problem is introduced. Then, a multiple-source shortest-path algorithm based heuristic for finding a suboptimal schedule of the communication steps for data staging is presented. A simulation study is provided, which evaluates the performance of the proposed heuristic. The results show the advantages of the proposed heuristic over two random based scheduling techniques. This research, based on the simplified static model, serves as a necessary step toward solving the more realistic and complicated version of the data staging problem involving dynamic scheduling, fault tolerance, and determining where to stage data.

Interconnection Networks for Parallel Computers

The sections in this article are 1 Network Topologies 2 Switching Techniques 3 Routing Techniques for Direct Networks 4 Switch Box Architectures 5 Conclusions

Bandwidth tracking in distributed heterogeneous networking environments

This work investigates bandwidth tracking algorithms in a version of a distributed heterogeneous data dissemination system called the Agile Information Control Environment (AICE). In this environment, the probability of setup or rejection of communication requests has to be derived without any updated knowledge of the actual topologies of the underlying networks. In this paper, a network learning algorithm is therefore introduced that is able to predict the setup/rejection of a communication request with a high accuracy of around 80% by tracking the average spare bandwidth of end-to-end communication channels. The learning algorithm uses elements of exponential growth coupled with a binary search. This enables the learner to quickly learn about changes in the network topology and traffic load. It is shown that the learner is able to predict request setup/rejection with a high accuracy without any information about underlying network topologies. To maximize the learner performance, an initial full mesh aggregation of the underlying network topology at system startup should be used.

Heuristics for scheduling prioritized data requests with deadlines in an overloaded distributed computing network

Gives an overview of research that the authors have conducted in the area of offline scheduling heuristics for communication requests in an overloaded network, where not all requests can be satisfied. Sites in the network request data items and each request has an associated deadline and priority. In a military situation, the data-staging problem involves positioning data for facilitating a faster access time when it is needed by programs that are to aid in decision-making. The work concentrates on solving a basic version of the data-staging problem in which all parameter values for the communication system and the data request information represent the best known information collected so far and stay fixed throughout the scheduling process. Three multiple-source shortest-path algorithm-based heuristics for finding a near-optimal schedule of the communication steps for staging the data are presented. Each heuristic is used with each of four cost criteria which have been developed. The performance of the proposed heuristics was evaluated and compared by simulations. The best heuristic was then combined with three variations of the best cost criterion; these variations consider the length of the path and the size of the data time requested. Further simulation studies were then performed. Also examined was the situation where two different versions of data items were available, with different sizes and different worths to the user. It is shown that the proposed heuristics perform very well with respect to an upper-bound measure.

Strategies for the Massively Parallel Simulation of Interconnection Networks

Methods for enhancing multistage interconnection network simulators running on massively parallel SIMD computers are presented. Aspects of parallel simulation of interconnection networks are discussed and different strategies of mapping the architecture of the network to be simulated onto the parallel machine are studied and compared. As case studies, two strategies of mapping synchronous multistage cube networks onto the MasPar MP-1 SIMD machine are explored and their implementations are compared. The methods result in an efficient simulator which can process 10^9 data packets in 40 minutes.