A heuristic method towards deadline-aware energy-efficient mapreduce scheduling problem in Hadoop YARN

Cluster Computing - The MapReduce (MR) scheduling is a prominent area of research to minimize energy consumption in the Hadoop framework in the era of green computing. Very few scheduling...     

A new approach to the job scheduling problem in computational grids

Job scheduling is one of the most challenging issues in Grid resource management that strongly affects the performance of the whole Grid environment. The major drawback of the existing Grid scheduling algorithms is that they are unable to adapt with the dynamicity of the resources and the network conditions. Furthermore, the network model that is used for resource information aggregation in most scheduling methods is centralized or semi-centralized. Therefore, these methods do not scale well as Grid size grows and do not perform well as the environmental conditions change with time. This paper proposes a learning automata-based job scheduling algorithm for Grids. In this method, the workload that is placed on each Grid node is proportional to its computational capacity and varies with time according to the Grid constraints. The performance of the proposed algorithm is evaluated through conducting several simulation experiments under different Grid scenarios. The obtained results are compared with those of several existing methods. Numerical results confirm the superiority of the proposed algorithm over the others in terms of makespan, flowtime, and load balancing.     

A novel energy-efficient scheduling model for multi-core systems

Cluster Computing - Multi-core systems has evolved enormously during the last decade with the improvement in the integration technology which makes it possible to house large number of transistors...     

Security-driven scheduling for data-intensive applications on grids

Security-sensitive applications that access and generate large data sets are emerging in various areas including bioinformatics and high energy physics. Data grids provide such data-intensive applications with a large virtual storage framework with unlimited power. However, conventional scheduling algorithms for data grids are unable to meet the security needs of data-intensive applications. In this paper we address the problem of scheduling data-intensive jobs on data grids subject to security constraints. Using a security- and data-aware technique, a dynamic scheduling strategy is proposed to improve quality of security for data-intensive applications running on data grids. To incorporate security into job scheduling, we introduce a new performance metric, degree of security deficiency, to quantitatively measure quality of security provided by a data grid. Results based on a real-world trace confirm that the proposed scheduling strategy significantly improves security and performance over four existing scheduling algorithms by up to 810% and 1478%, respectively.

A comparison of utility-oriented algorithms for scheduling parallel tasks in multi-cluster grid

Scheduling parallel tasks in multi-cluster grid can be seen as two interdependent problems: cluster allocation and scheduling parallel task on the allocated cluster. In this paper both rigid and moldable parallel tasks are considered. We propose a theoretical model of utility-oriented parallel task scheduling in multi-cluster grid with advance reservations. On the basis of the model we present an approximation algorithm, a repair strategy based genetic algorithm and greedy heuristics MaxMax, T-Sufferage and R-Sufferage to solve the two interdependent problems. We compare the performance of these algorithms in aspect of utility optimality and timing results. Simulation results show on average the (1+α)-approximation algorithm achieves the best trade-off between utility optimality and timing. Genetic algorithm could achieve better utility than greedy heuristics and approximate algorithm at expensive time cost. Greedy heuristics do not perform equally well when adapted to different utility functions while the approximation algorithm shows its intrinsic stable performance.     

HSGA: a hybrid heuristic algorithm for workflow scheduling in cloud systems

In heterogeneous distributed computing systems like cloud computing, the problem of mapping tasks to resources is a major issue which can have much impact on system performance. For some reasons such as heterogeneous and dynamic features and the dependencies among requests, task scheduling is known to be a NP-complete problem. In this paper, we proposed a hybrid heuristic method (HSGA) to find a suitable scheduling for workflow graph, based on genetic algorithm in order to obtain the response quickly moreover optimizes makespan, load balancing on resources and speedup ratio. At first, the HSGA algorithm makes tasks prioritization in complex graph considering their impact on others, based on graph topology. This technique is efficient to reduction of completion time of application. Then, it merges Best-Fit and Round Robin methods to make an optimal initial population to obtain a good solution quickly, and apply some suitable operations such as mutation to control and lead the algorithm to optimized solution. This algorithm evaluates the solutions by considering efficient parameters in cloud environment. Finally, the proposed algorithm presents the better results with increasing number of tasks in application graph in contrast with other studied algorithms.     

Task scheduling on computational Grids using Gravitational Search Algorithm

Grid computing uses distributed interconnected computers and resources collectively to achieve higher performance computing and resource sharing. Task scheduling is one of the core steps to efficiently exploit the capabilities of Grid environment. Recently, heuristic algorithms have been successfully applied to solve task scheduling on computational Grids. In this paper, Gravitational Search Algorithm (GSA), as one of the latest population-based metaheuristic algorithms, is used for task scheduling on computational Grids. The proposed method employs GSA to find the best solution with the minimum makespan and flowtime. We evaluate this approach with Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) method. The results demonstrate that the benefit of the GSA is its speed of convergence and the capability to obtain feasible schedules.     

An efficient heuristic for scheduling batches of parts in a flexible flow system

Flexible manufacturing systems (FMSs) are a class of automated systems that can be used to improve productivity in batch manufacturing. Four stages of decision making have been defined for an FMS—the design, planning, scheduling, and control stages. This research focuses on the planning stage, and specifically in the area of scheduling batches of parts through the system. The literature to date on the FMS planning stage has mostly focused on the machine grouping, tool loading, and parttype selection problems. Our research carries the literature a step further by addressing the problem of scheduling batches of parts. Due to the use of serial-access material-handling systems in many FMSs, the batch-scheduling problem is modeled for a flexible flow system (FFS). This model explicitly accounts for setup times between batches that are dependent on their processing sequence. A heuristic procedure is developed for this batch-scheduling problem—the Maximum Savings (MS) heuristic. The MS heuristic is based upon the savings in time associated with a particular sequence and selecting the one with the maximum savings. It uses a two-phase method, with the savings being calculated in phase I, while a branch-and-bound procedure is employed to seek the best heuristic solution in phase II. Extensive computational results are provided for a wide variety of problems. The results show that the MS heuristic provides good-quality solutions.     

Energy aware scheduling of deadline-constrained tasks in cloud computing

Energy efficiency is the predominant issue which troubles the modern ICT industry. The ever-increasing ICT innovations and services have exponentially added to the energy demands and this proliferated the urgency of fostering the awareness for development of energy efficiency mechanisms. But for a successful and effective accomplishment of such mechanisms, the support of underlying ICT platform is significant. Eventually, Cloud computing has gained attention and has emerged as a panacea to beat the energy consumption issues. This paper scrutinizes the importance of multicore processors, virtualization and consolidation techniques for achieving energy efficiency in Cloud computing. It proposes Green Cloud Scheduling Model (GCSM) that exploits the heterogeneity of tasks and resources with the help of a scheduler unit which allocates and schedules deadline-constrained tasks delimited to only energy conscious nodes. GCSM makes energy-aware task allocation decisions dynamically and aims to prevent performance degradation and achieves desired QoS. The evaluation and comparative analysis of the proposed model with two other techniques is done by setting up a Cloud environment. The results indicate that GCSM achieves 71 % of energy savings and high performance in terms of deadline fulfillment.     

Energy and cost trade-off for computational tasks offloading in mobile multi-tenant clouds

Cluster Computing - Mobile cloud computing augments smart-phones with computation capabilities by offloading computations to the cloud. Recent works only consider the energy savings of mobile...     

Abundance of correctly folded RNA motifs in sequence space, calculated on computational grids

Although functional RNA molecules are known to be biased in overall composition, the effects of background composition on the probability of finding a particular active site by chance has received little attention. The probability of finding a particular motif has important implications both for understanding the distribution of functional RNAs in ancient and modern organisms with varying genome compositions and for tuning SELEX pools to optimize the chance of finding specific functions. Here we develop a new method for calculating the probability of finding a modular motif containing base-paired regions, and use a computational grid to fold several hundred million random RNA sequences containing the core elements of the isoleucine aptamer and the hammerhead ribozyme to estimate the probability that a sequence containing these structural elements will fold correctly when isolated from background sequences of different compositions. We find that the two motifs are most likely to be found in distinct regions of compositional space, and that the regions of greatest abundance are influenced by the probability of finding the conserved bases, finding the flanking helices, and folding, in that order of importance. Additionally, we can refine our estimates of the number of random sequences required for a 50% probability of finding an example of each site in unbiased random pools of length 100 to 4.1 × 10⁹ for the isoleucine aptamer and 1.6 × 10¹⁰ for the hammerhead ribozyme. These figures are consistent with the facile recovery of these motifs from SELEX experiments.     

A heterogeneity-aware approach to load balancing of computational tasks: a theoretical and simulation study

One of the distinct characteristics of computing platforms shared by multiple users such as a cluster and a computational grid is heterogeneity on each computer and/or among computers. Temporal heterogeneity refers to variation, along the time dimension, of computing power available for a task on a computer, and spatial heterogeneity represents the variation among computers. In minimizing the average parallel execution time of a target task on a spatially heterogeneous computing system, it is not optimal to distribute the target task linearly proportional to the average computing powers available on computers. In this paper, effects of the temporal and spatial heterogeneity on performance of a target task have been analyzed in terms of the mean and standard deviation of parallel execution time. Based on the analysis results, an approach to load balancing for minimizing the average parallel execution time of a target task is described. The proposed approach whose validity has been verified through simulation considers temporal and spatial heterogeneities in addition to the average computing power on each computer.

A heuristic algorithm for identical parallel machine scheduling: splitting jobs,sequence-dependent setup times,and limited setup operators

Flexible Services and Manufacturing Journal - We examine a parallel machine scheduling problem with a job splitting property, sequence-dependent setup times, and limited setup operators, for...     

A heuristic managing errors for DNA sequencing

MOTIVATION: A new heuristic algorithm for solving DNA sequencing by hybridization problem with positive and negative errors. RESULTS: A heuristic algorithm providing better solutions than algorithms known from the literature based on tabu search method.     

A comparative analysis of dynamic grids vs. virtual grids using the A3pviGrid framework

With the proliferation of Quad/Multi-core micro-processors in mainstream platforms such as desktops and workstations; a large number of unused CPU cycles can be utilized for running virtual machines (VMs) as dynamic nodes in distributed environments. Grid services and its service oriented business broker now termed cloud computing could deploy image based virtualization platforms enabling agent based resource management and dynamic fault management. In this paper we present an efficient way of utilizing heterogeneous virtual machines on idle desktops as an environment for consumption of high performance grid services. Spurious and exponential increases in the size of the datasets are constant concerns in medical and pharmaceutical industries due to the constant discovery and publication of large sequence databases. Traditional algorithms are not modeled at handing large data sizes under sudden and dynamic changes in the execution environment as previously discussed. This research was undertaken to compare our previous results with running the same test dataset with that of a virtual Grid platform using virtual machines (Virtualization). The implemented architecture, A3pviGrid utilizes game theoretic optimization and agent based team formation (Coalition) algorithms to improve upon scalability with respect to team formation. Due to the dynamic nature of distributed systems (as discussed in our previous work) all interactions were made local within a team transparently. This paper is a proof of concept of an experimental mini-Grid test-bed compared to running the platform on local virtual machines on a local test cluster. This was done to give every agent its own execution platform enabling anonymity and better control of the dynamic environmental parameters. We also analyze performance and scalability of Blast in a multiple virtual node setup and present our findings. This paper is an extension of our previous research on improving the BLAST application framework using dynamic Grids on virtualization platforms such as the virtual box.     

A heuristic test for homogeneity in species populations

Summary This paper elaborates on the notion of homogeneity in plant populations and its quantification. It describes a method for testing homogeneity without the need to make assumptions about generalized distributions. The implementation of the method is illustrated through actual numerical examples.This paper presents results from a study of plant populations for which a National Research Council of Canade grant has been received (L. Orlóci).