Cooperation in multi‐organization scheduling

Pascual, Fanny; Rządca, Krzysztof; Trystram, Denis

doi:10.1002/cpe.1378

Cited by 22 publications

(20 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We just mention main results, without details because they are not addressed in our paper. In [29], a related clairvoyant problem was addressed. The authors model an offline system consisting of N clusters with exactly m identical processors each, and propose an algorithm with a guaranteed worst-case performance ratio on the global makespan equal to 4.…”

Section: Theoretical Resultsmentioning

confidence: 99%

Job Allocation Strategies with User Run Time Estimates for Online Scheduling in Hierarchical Grids

et al. 2011

View full text Add to dashboard Cite

We address non-preemptive nonclairvoyant online scheduling of parallel jobs on a Grid. We consider a Grid scheduling model with two stages. At the first stage, jobs are allocated to a suitable Grid site, while at the second stage, local scheduling is independently applied to each site. We analyze allocation strategies depending on the type and amount of information they require. We conduct a comprehensive performance evaluation study using simulation and demonstrate that our strategies perform well with respect to several metrics that reflect both user-and system-centric goals. Unfortunately, user run time estimates and information on local schedules does not help to significantly improve the outcome of the allocation strategies. When examining the overall Grid performance based on real data, we determined that an appropriate distribution of job processor requirements over the Grid has a higher performance than an allocation of jobs based on user run time estimates and information on local schedules. In general, our experiments showed that rather simple schedulers with minimal information requirements can provide a good performance.

show abstract

Section: Theoretical Resultsmentioning

confidence: 99%

Job Allocation Strategies with User Run Time Estimates for Online Scheduling in Hierarchical Grids

et al. 2011

View full text Add to dashboard Cite

show abstract

“…They presented a 3-approximation for the maximum completion time (makespan) when all jobs are known in advance (which may correspond, for instance, to a batch) and a 5-approximation for the on-line case. Pascual et al [15], [16] extended the problem for the case where resources and jobs are aggregated in independent organizations that have a local objective for their jobs besides the global makespan. Their main contribution is the analysis of a centralized 3-approximation algorithm that always incite these organizations to cooperate.…”

Section: B Related Workmentioning

confidence: 99%

“…The problem was first introduced by Pascual et al [15] and models scheduling on grid computing systems, with particular interest on the performance objectives of each individual organization that compounds the grid computing system.…”

Section: A the Multi-organization Scheduling Problemmentioning

confidence: 99%

See 1 more Smart Citation

Coordination mechanisms for selfish multi-organization scheduling

Cohen¹,

Cordeiro²,

Trystram³

et al. 2011

2011 18th International Conference on High Performance Computing

Self Cite

View full text Add to dashboard Cite

We conduct a game theoretic analysis on the problem of scheduling jobs on computing platforms composed of several independent and selfish organizations, known as the MultiOrganization Scheduling Problem (MOSP). Each organization shares resources and jobs with others, expecting to decrease the makespan of its own jobs.We modeled MOSP as a non-cooperative game where each agent is responsible for assigning all jobs belonging to a particular organization to the available processors. The local scheduling of these jobs is defined by coordination mechanisms that first prioritize local jobs and then schedule the jobs from others according to some given priority. When different priorities are given individually to the jobs -like in classical scheduling algorithms such as LPT or SPT -then no pure -approximate equilibrium is possible for values of less than 2. We also prove that even deciding whether a given instance admits or not a pure Nash equilibrium is co-NP hard. When these priorities are given to entire organizations, we show the existence of an algorithm that always computes a pure ρ-approximate equilibrium using any ρ-approximation list scheduling algorithm. Finally, we prove that the price of anarchy of the MOSP game using this mechanism is asymptotically bounded by 2.

show abstract

“…Various resource-efficient and economically beneficent allocation techniques and methodologies have already been proposed to address these issues. These methodologies include game theoretical approaches [3], [45], stochastic programming [17], bio-inspired mechanisms [20], [41], auction-based algorithms [25], [36], [46] and agent-based approaches [21], [40], [44]. Most of these works perform QoS service selection and resource allocation based on some service performance indicators and economic indicators, including wait time optimization, utilization maximization and economic wastage minimization.…”

Section: Cloud Markets In Researchmentioning

confidence: 99%

Maximizing Liquidity in Cloud Markets through Standardization of Computational Resources

Breskovic

Brandić

Altmann

2013

2013 IEEE Seventh International Symposium on Service-Oriented System Engineering

View full text Add to dashboard Cite

Abstract-Low liquidity in cloud markets can result in market instability and inefficiency, preventing the successful implementation of ubiquitous computing on demand. To circumvent this issue, it has been suggested to channel demand and supply into a limited number of standardized services. These standardized services can even be automatically adapted to user requirements with the goal of continuously improving market performance. In this paper, we focus on answering how many standardized services should be placed in the market. This work is based on a new definition of liquidity for cloud resources, which in turn has been derived from liquidity definitions of financial markets. Using a simulation framework, we evaluate our method for estimating the optimal quantity of standardized services with respect to market liquidity and demonstrate the benefits of this approach in terms of increase in market efficiency and decrease in users' cost of participation in the market. The methods presented in this paper have the potential to be applied in other electronic markets as well.Keywords-Service level agreement, electronic markets, cloud economics, autonomic computing, standardized goods, market modeling, market liquidity, IT services I. INTRODUCTION Common to utility computing, cloud computing and grid computing is that they offer computational resources (e.g., software, hardware and computing platform services) in a manner similar to utilities, such as water, electricity and telephony, without regard to where the services are hosted or how they are delivered [4], [15]. However, the current cloud market is fragmented and static, hindering the paradigm's ability to fulfill its promise of ubiquitous computing on demand and as a commodity. In order to address this issue, electronic markets for trading and/or allocating computational resources (i.e., grid and cloud services) have been proposed [1], [38].In cloud markets, computational services are described by service parameters and quality of service (QoS) objectives. Each setting of a service, i.e., configuration of parameters, desired values of QoS objectives, and combinations of these elements, defines a new service (also called "computational resource"). Due to a great number of available settings, markets suffer from a vast diversity and heterogeneity of computational resources.In addition to this, computational resources are characterized by a high dynamism and a high fungibility. Resource dynamism is a result of a large resource variability, a dynamic user base, diverse user behavior in the market, and the

show abstract

Cooperation in multi‐organization scheduling

Cited by 22 publications

References 35 publications

Job Allocation Strategies with User Run Time Estimates for Online Scheduling in Hierarchical Grids

Job Allocation Strategies with User Run Time Estimates for Online Scheduling in Hierarchical Grids

Coordination mechanisms for selfish multi-organization scheduling

Maximizing Liquidity in Cloud Markets through Standardization of Computational Resources

Contact Info

Product

Resources

About