Johnatan E. Pecero

A survey on resource allocation in high performance distributed computing systems

Classification of high performance computing (HPC) systems is provided.Current HPC paradigms and industrial application suites are discussed.State of the art in HPC resource allocation is reported.Hardware and software solutions are discussed for optimized HPC systems. An efficient resource allocation is a fundamental requirement in high performance computing (HPC) systems. Many projects are dedicated to large-scale distributed computing systems that have designed and developed resource allocation mechanisms with a variety of architectures and services. In our study, through analysis, a comprehensive survey for describing resource allocation in various HPCs is reported. The aim of the work is to aggregate under a joint framework, the existing solutions for HPC to provide a thorough analysis and characteristics of the resource management and allocation strategies. Resource allocation mechanisms and strategies play a vital role towards the performance improvement of all the HPCs classifications. Therefore, a comprehensive discussion of widely used resource allocation strategies deployed in HPC environment is required, which is one of the motivations of this survey. Moreover, we have classified the HPC systems into three broad categories, namely: (a) cluster, (b) grid, and (c) cloud systems and define the characteristics of each class by extracting sets of common attributes. All of the aforementioned systems are cataloged into pure software and hybrid/hardware solutions. The system classification is used to identify approaches followed by the implementation of existing resource allocation strategies that are widely presented in the literature.

Energy-Aware Scheduling on Multicore Heterogeneous Grid Computing Systems

We address a multicriteria non-preemptive energy-aware scheduling problem for computational Grid systems. This work introduces a new formulation of the scheduling problem for multicore heterogeneous computational Grid systems in which the minimization of the energy consumption, along with the makespan metric, is considered. We adopt a two-level model, in which a meta-broker agent (level 1) receives all user tasks and schedules them on the available resources, belonging to different local providers (level 2). The computing capacity and energy consumption of resources are taken from real multi-core processors from the main current vendors. Twenty novel list scheduling methods for the problem are proposed, and a comparative analysis of all of them over a large set of problem instances is presented. Additionally, a scalability study is performed in order to analyze the contribution of the best new bi-objective list scheduling heuristics when the problem dimension grows. We conclude after the experimental analysis that accurate trade-off schedules are computed by using the new proposed methods.

A two-phase heuristic for the energy-efficient scheduling of independent tasks on computational grids

The sensitivity analysis of a Cellular Genetic Algorithm (CGA) with local search is used to design a new and faster heuristic for the problem of mapping independent tasks to a distributed system (such as a computer cluster or grid) in order to minimize makespan (the time when the last task finishes). The proposed heuristic improves the previously known Min-Min heuristic. Moreover, the heuristic finds mappings of similar quality to the original CGA but in a significantly reduced runtime (1,000 faster). The proposed heuristic is evaluated across twelve different classes of scheduling instances. In addition, a proof of the energy-efficiency of the algorithm is provided. This convergence study suggests how additional energy reduction can be achieved by inserting low power computing nodes to the distributed computer system. Simulation results show that this approach reduces both energy consumption and makespan.

CA-DAG: Modeling Communication-Aware Applications for Scheduling in Cloud Computing

This paper addresses performance issues of resource allocation in cloud computing. We review requirements of different cloud applications and identify the need of considering communication processes explicitly and equally to the computing tasks. Following this observation, we propose a new communication-aware model of cloud computing applications, called CA-DAG. This model is based on Directed Acyclic Graphs that in addition to computing vertices include separate vertices to represent communications. Such a representation allows making separate resource allocation decisions: assigning processors to handle computing jobs, and network resources for information transmissions. The proposed CA-DAG model creates space for optimization of a number of existing solutions to resource allocation and for developing novel scheduling schemes of improved efficiency.

Online Bi-Objective Scheduling for IaaS Clouds Ensuring Quality of Service

This paper focuses on a bi-objective experimental evaluation of online scheduling in the Infrastructure as a Service model of Cloud computing regarding income and power consumption objectives. In this model, customers have the choice between different service levels. Each service level is associated with a price per unit of job execution time, and a slack factor that determines the maximal time span to deliver the requested amount of computing resources. The system, via the scheduling algorithms, is responsible to guarantee the corresponding quality of service for all accepted jobs. Since we do not consider any optimistic scheduling approach, a job cannot be accepted if its service guarantee will not be observed assuming that all accepted jobs receive the requested resources. In this article, we analyze several scheduling algorithms with different cloud configurations and workloads, considering the maximization of the provider income and minimization of the total power consumption of a schedule. We distinguish algorithms depending on the type and amount of information they require: knowledge free, energy-aware, and speed-aware. First, to provide effective guidance in choosing a good strategy, we present a joint analysis of two conflicting goals based on the degradation in performance. The study addresses the behavior of each strategy under each metric. We assess the performance of different scheduling algorithms by determining a set of non-dominated solutions that approximate the Pareto optimal set. We use a set coverage metric to compare the scheduling algorithms in terms of Pareto dominance. We claim that a rather simple scheduling approach can provide the best energy and income trade-offs. This scheduling algorithm performs well in different scenarios with a variety of workloads and cloud configurations.

Adaptive energy efficient scheduling in Peer-to-Peer desktop grids

Abstract We address non-preemptive scheduling problems on heterogeneous P2P grids, where resources are changing over time, and scheduling decisions are free from information of application characteristics. We consider a scheduling with task replications to overcome possible bad resource allocation in presence of uncertainty, and ensure good performance. We analyze the energy consumption of job allocation strategies exploring the replication thresholds, and dynamic component deactivation. The main idea of our approach is to set replication thresholds, and dynamically adapt them to cope with different objective preferences, workloads, and Grid properties. We compare three groups of strategies: knowledge-free, speed-aware, and power-aware. In order to provide good performance and minimize energy consumption, first, we perform a joint analysis of two metrics considering their degradation in performance. Then, we provide two-objective optimization analysis that is not restricted to find a unique solution, but the Pareto optimal set. Based on these results, we use a Set Coverage metric for assessing the performance of our strategies and compare twenty algorithms in terms of Pareto dominance. A case study is given, and corresponding results indicate that two replicas for knowledge-free algorithms, and one replica for speed-aware algorithms provide the best energy and performance trade-offs in the scheduling. They perform well in different scenarios with a variety of workloads and grid configurations.

Multi-objective evolutionary algorithms for energy-aware scheduling on distributed computing systems

HighlightsHeterogeneous, energy-aware precedence constrained (DAG) scheduling problem.Three multi-objective algorithms schemas are adapted: MOCell, NSGAII and IBEA.A representation with corresponding mutations and grouping crossover operators.Experimentation on a diversified and large set of real and synthetic applications.MOCell schema is the most versatile and the best performing one. The ongoing increase of energy consumption by IT infrastructures forces data center managers to find innovative ways to improve energy efficiency. The latter is also a focal point for different branches of computer science due to its financial, ecological, political, and technical consequences. One of the answers is given by scheduling combined with dynamic voltage scaling technique to optimize the energy consumption. The way of reasoning is based on the link between current semiconductor technologies and energy state management of processors, where sacrificing the performance can save energy.This paper is devoted to investigate and solve the multi-objective precedence constrained application scheduling problem on a distributed computing system, and it has two main aims: the creation of general algorithms to solve the problem and the examination of the problem by means of the thorough analysis of the results returned by the algorithms.The first aim was achieved in two steps: adaptation of state-of-the-art multi-objective evolutionary algorithms by designing new operators and their validation in terms of performance and energy. The second aim was accomplished by performing an extensive number of algorithms executions on a large and diverse benchmark and the further analysis of performance among the proposed algorithms. Finally, the study proves the validity of the proposed method, points out the best-compared multi-objective algorithm schema, and the most important factors for the algorithms performance.

A Cellular Genetic Algorithm for scheduling applications and energy-aware communication optimization

In modern parallel and distributed systems, inter-processor communications are a crucial factor of performance. The time for exchanging data is usually larger than that for computing elementary operations. Consequently, these communications slow down the execution of the application scheduled on the computing platform. Accounting for these communications is essential for attaining efficient hardware and software utilization. Moreover, energy dissipation due to the transfer of data between processing elements has become a major concern. Therefore, in this paper we develop an energy-aware static algorithm, which intrinsically optimizes the energy consumption due to the transfer of data in a distributed system. This is achieved by properly allocating and scheduling the tasks that constitute the applications on the processing elements, minimizing inter-processor communications. The proposed algorithm is a new Cellular Genetic Algorithm based on task clustering techniques. That is, the genetic operators work considering groups of tasks instead of applying them directly on the tasks. Simulation results showed that this algorithm is very compelling in terms of application completion time, inter-processor communication and energy communication dissipation.

Energy-aware fast scheduling heuristics in heterogeneous computing systems

In heterogeneous computing systems it is crucial to schedule tasks in a manner that exploits the heterogeneity of the resources and applications to optimize systems performance. Moreover, the energy efficiency in these systems is of a great interest due to different concerns such as operational costs and environmental issues associated to carbon emissions. In this paper, we present a series of original low complexity energy efficient algorithms for scheduling. The main idea is to map a task to the machine that executes it fastest while the energy consumption is minimum. On the practical side, the set of experimental results showed that the proposed heuristics perform as efficiently as related approaches, demonstrating their applicability for the considered problem and its good scalability.

A two-phase heuristic for the scheduling of independent tasks on computational grids

The sensitivity analysis of a cellular genetic algorithm with local search is used to design a new and simpler heuristic for the problem of scheduling independent tasks. The proposed heuristic improves the previously known Min-Min heuristic. Moreover, it provides schedules of similar quality to the reference cellular genetic algorithm in a significantly reduced runtime. This heuristic is evaluated across twelve different classes of scheduling instances.