Cesar O. Diaz

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.

Scalable and Energy-Efficient Scheduling Techniques for Large-Scale Systems

The scalability of a computing system can be identified by at least three components: (a) size, (b) geographical distribution, and (c) administrative constraints. Newer paradigms, such as clouds, grids, and clusters bring in more parameters to the aforementioned list, namely heterogeneity, energy consumption, and transparency. To optimize the performance of a computing system, it is manner that exploits heterogeneity and is scalable. Moreover, newer systems also demand energy efficiency as an integral part of schedulers. In this paper, we evaluate the behavior of low complexity energy-efficient algorithms for scheduling. The set of experimental results showed that the evaluated heuristics perform as efficiently as related approaches, demonstrating their applicability and scalability for the considered problem.

Scalable, low complexity, and fast greedy scheduling heuristics for highly heterogeneous distributed computing systems

For heterogeneous distributed computing systems, important design issues are scalability and system optimization. Given such systems, it is crucial to develop low computational complexity algorithms to schedule tasks in a manner that exploits the heterogeneity of the resources and applications. In this paper, we report and evaluate three scalable, and fast scheduling heuristics for highly heterogeneous distributed computing systems. We conduct a comprehensive performance evaluation study using simulation. The benchmarking outlines the performance of the schedulers, representing scalability, makespan, flowtime, computational complexity, and memory utilization. The set of experimental results shows that our heuristics perform as good as the traditional approaches, for makespan and flowtime, while featuring lower complexity, lower running time, and lower used memory. The experimental results also detail the various scenarios under which certain algorithms excel and fail.

Green flexible opportunistic computing with task consolidation and virtualization

Energy efficiency and high computing power are basic design considerations across modern-day computing solutions due to different concerns such as system performance, operational cost, and environmental issues. Desktop Grid and Volunteer Computing System (DGVCS) so called opportunistic infrastructures offer computational power at low cost focused on harvesting idle computing cycles of existing commodity computing resources. Other than allowing to customize the end user offer, virtualization is considered as one key techniques to reduce energy consumption in large-scale systems and contributes to the scalability of the system. This paper presents an energy efficient approach for opportunistic infrastructures based on task consolidation and customization of virtual machines. The experimental results with single desktops and complete computer rooms show that virtualization significantly improves the energy efficiency of opportunistic grids compared with dedicated computing systems without disturbing the end-user.

Impact of Voltage Levels Number for Energy-Aware Bi-objective DAG Scheduling for Multi-processors Systems

The paper investigates the influence of Dynamic Voltage Frequency Scaling for bi-objective (makespan and energy consumption) Directed Acyclic task Graph scheduling on heterogeneous multi-processor platform. The proposed resolution method of solving conflicting criteria relies on Multi-Objective Evolutionary Algorithms. Two voltage frequency approaches are compared: one using only 2 levels (minimal and maximal) and the other one using a larger number. The approaches are benchmarked on applications with Laplace transformation and Gaussian elimination structures using the NSGAII algorithm. The results show that while the Two-level approach generates more discriminated solutions on the Pareto front, the solutions are of a lower quality than in the Multi-level approach.

Green Flexible Opportunistic Computing with Virtualization

Energy efficiency and High computing power are basic design considerations across modern-day computing solutions due to different concerns such as system performance, operational cost, and environmental issues. Opportunistic grid infrastructures offer computational power at low cost focused on harvesting idle computing cycles of existing commodity computing resources. Other than allowing to customize the end user offer, virtualization is considered as one key techniques to reduce energy consumption in large-scale systems and contributes to the scalability of the system. This paper presents an energy efficient approach for opportunistic grids based on virtualization. The experimental results show that virtualization significantly improves the energy efficiency of opportunistic grids compared with dedicated computing systems without disturbing the end-user.

Determining the Real Capacity of a Desktop Cloud

Computer laboratories at Universities are underutilized most of the time [1]. Having an averaged measure of its computing resources usage would allow researchers to harvest the capacity available by deploying opportunistic infrastructures, that is, infrastructures mostly supported by idle computing resources which run in parallel to tasks performed by the resource owner (end-user). In this paper we measure such usage in terms of CPU and RAM. The metrics were obtained by using the SIGAR library on 70 desktops belonging to two independent laboratories during the three busiest weeks in the semester. We found that the averaged usage of CPU is less than 5 % while RAM is around 25 %. The results show that in terms of the amount of floating point operations per second (FLOPS) there is a capacity of 24 GFLOPS that can be effectively harvest by deploying opportunistic infrastructures to support e-Science without affecting the performance perceived by end-users and avoiding underutilization and the acquisition of new hardware.

Performance Evaluation of an IaaS Opportunistic Cloud Computing

This poster shows the performance evaluation of UnaCloud Opportunistic Computing IaaS. We analyze from an HPC perspective, two virtualization frameworks Virtual Box and VMware ESXi and compare them over this particular opportunistic cloud environment. The benchmarks consist of two set of tests, High Performance Linpack and IOzone, that examine the performance and the Input/Output response. The purpose of the experiments is to evaluate the behavior of the different virtual environments over an opportunistic cloud environment and investigate how these are affected by different percentage of end-users. The results show a better performance for Virtual Box than VMware and the other way around for I/O response. Nevertheless, the experiments shows that VBox have more robustness than VMware.

Energy-aware VM Allocation on an Opportunistic Cloud Infrastructure

UnaCloud is an opportunistic based cloud infrastructure (IaaS) that allows to access on-demand computing capabilities using commodity desktops. Although UnaCloud maximizes the use of idle resources to deploy virtual machines, it does not use energy-efficient resource allocation algorithms. In this paper, we design and develop different energy-aware algorithms to operate in an energy-efficient way and at the same time to guarantee the performance of the UnaCloud users. Performance tests with different algorithms and scenarios using real trace workloads from UnaCloud, show how different policies can change the energy consumption patterns and reduce the energy consumption in the opportunistic cloud infrastructure. The results show that some algorithms can reduce the energy-consumption power up to 30% over the percentage earned by the opportunistic environment.

Energy Savings on a Cloud-Based Opportunistic Infrastructure

In this paper, we address energy savings on a Cloud-based opportunistic infrastructure. The infrastructure implements opportunistic design concepts to provide basic services, such as virtual CPUs, RAM and Disk while profiting from unused capabilities of desktop computer laboratories in a non-intrusive way.