Dynamic scheduling for heterogeneous Desktop Grids

Al-Azzoni, Issam; Down, Douglas G.

doi:10.1109/grid.2008.4662792

Cited by 11 publications

(9 citation statements)

References 15 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A recent work that shares some similarities with our work presents PROPHET, a goaloriented provisioning infrastructure that tunes the WSC to satisfy the needs of particular end users [33]. The research work that is closest to ours discusses a scheduling policy which uses a linear programming problem that maximizes system capacity to map an application across a desktop grid [4]. This work focuses on distributed desktop computers and does not consider the interaction between microarchitectural and co-location heterogeneity.…”

Section: Related Workmentioning

confidence: 99%

Whare-map

Jason

Tang

2013

SIGARCH Comput. Archit. News

View full text Add to dashboard Cite

Modern "warehouse scale computers" (WSCs) continue to be embraced as homogeneous computing platforms. However, due to frequent machine replacements and upgrades, modern WSCs are in fact composed of diverse commodity microarchitectures and machine configurations. Yet, current WSCs are architected with the assumption of homogeneity, leaving a potentially significant performance opportunity unexplored.In this paper, we expose and quantify the performance impact of the "homogeneity assumption" for modern production WSCs using industry-strength large-scale web-service workloads. In addition, we argue for, and evaluate the benefits of, a heterogeneity-aware WSC using commercial web-service production workloads including Google's websearch. We also identify key factors impacting the available performance opportunity when exploiting heterogeneity and introduce a new metric, opportunity factor, to quantify an application's sensitivity to the heterogeneity in a given WSC. To exploit heterogeneity in "homogeneous" WSCs, we propose "Whare-Map," the WSC Heterogeneity Aware Mapper that leverages already in-place continuous profiling subsystems found in production environments. When employing "Whare-Map", we observe a cluster-wide performance improvement of 15% on average over heterogeneityoblivious job placement and up to an 80% improvement for web-service applications that are particularly sensitive to heterogeneity.

show abstract

Section: Related Workmentioning

confidence: 99%

Whare-map

Jason

Tang

2013

SIGARCH Comput. Archit. News

View full text Add to dashboard Cite

show abstract

“…The problem of scheduling concurrent multiple BoT applications using centralized or hierarchical frameworks has recently been proposed [6], [7]. Benoit et al [6] addressed the scheduling problem on heterogeneous master-worker frameworks with Linear Programming (LP) model.…”

Section: Literature Reviewmentioning

confidence: 99%

“…By expressing the problem in canonical form with appropriate constraints, the scheduler can discover a schedule in polynomial-time to make sure each application is executed fairly. Al-Azzoni and Down [7] also transform the problem to a linear program aiming to maximize the system utilization. In this model, resources in the network are assumed to be time-shared where the portion of processor cycle donated to public tasks can vary with time.…”

Section: Literature Reviewmentioning

confidence: 99%

“…The Bag-of-Tasks (BoT) application model [5], i.e., an embarrassingly parallel application composed of a collection of independent and identical tasks, is often considered as a suitable model for desktop grids [6], [7]. Scheduling concurrent BoT applications is challenging because any delay in completion of any one of the constituent tasks of the bag will affect the overall bag turnaround time.…”

Section: Introductionmentioning

confidence: 99%

“…Scheduling concurrent BoT applications is challenging because any delay in completion of any one of the constituent tasks of the bag will affect the overall bag turnaround time. To the best of our knowledge, researchers have only addressed scheduling concurrent bags in centralized desktop grids [6], [7], [8], traditional grid computing [9], [10], [11] or enterprise desktop grids [12], [13]. However the problem of scheduling BoT applications in dynamic and volatile Internet-scale volunteer computing has not been considered.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Bag-of-Tasks applications scheduling on volunteer desktop grids with adaptive information dissemination

Kwan

Jogesh

2010

IEEE Local Computer Network Conference

View full text Add to dashboard Cite

In this paper, we explore the problem of scheduling multiple concurrent Bag-of-Tasks (BoT) applications in unstructured peer-to-peer (P2P) volunteer desktop grids built around a super-peer framework. Resource availability information is exchanged among the super-peers using a light-weight thresholddriven gossip protocol with the aim of minimizing the resource discovery overhead. We present a comprehensive study of the properties of fixed threshold and adaptive threshold based gossiping mechanisms using detailed simulation experiments.Our results indicate that while gossiping with fixed threshold reflects resource usage effectively in a scalable network, an adaptive gossip threshold mechanism based on a simple feedback mechanism considering system state adapts much better to a volatile environment with small latency, reduced overhead and yields better system performance compared to fixed threshold protocols.

show abstract

Task scheduling scheme based on resource clustering in desktop grids

Gil

Kim

Lee

2013

Int J Communication

View full text Add to dashboard Cite

Desktop grids are platforms for grid computing that incorporate desktop resources into a grid infrastructure. The purpose of this computing paradigm is to process a massive computational tasks by exploiting the donated resources connected over the Internet. In desktop grids, it is important to guarantee fast turnaround time in the presence of dynamic properties, such as volatility and heterogeneity. To achieve this objective, we propose a task scheduling scheme based on resource clustering that can selectively allocate tasks to those resources that are most suitable for the current situation of a desktop grid environment. As a classifier of resources, the k-means clustering algorithm is introduced to classify resources according to their own task execution availability and result-return probability. The experimental results show that our scheduling scheme is more efficient than existing scheduling schemes with respect to reducing both the turnaround time and the quantity of resources consumed. TASK SCHEDULING SCHEME BASED ON RESOURCE CLUSTERING IN DESKTOP GRIDS 919 grid computing [1], each resources in a desktop grid can freely join and leave during task execution. Accordingly, task execution can be volatile at any time, and thus task failures inevitably occur. Desktop grids are also based on heterogeneous resources from different types of CPUs, OSs, memory capacity, and so on [16]. Task completion time can be vary greatly depending on the performance of resources. Therefore, task scheduling in desktop grids must positively consider the dynamic properties that arise from these volatility and heterogeneity.Task scheduling is the process of assigning tasks to the most suitable resources [17][18][19]. Unlike traditional approach in grids, task scheduling in desktop grids should consider the type of resources, resource trust, execution failures, applications, and so on [20]. Tasks can be allocated to resources either statically or dynamically [17]. When tasks are statically allocated, all of the information needed for scheduling, such as task execution time, resource performance, and the number of resources participating, should be known in advance. Thus, tasks can be deterministically allocated to resources before task execution is started. On the other hand, dynamic task scheduling is based on a continuous request of incoming tasks, and thus a decision for task allocation is postponed until run time; tasks should be allocated to resources, taking into account the current environmental information such as available resources, the number of uncompleted tasks, the network situation, and so on.Static task scheduling is not suitable for desktop grids due to the dynamic characteristics caused by volatility and heterogeneity. Thus, dynamic task scheduling using the current environmental information should be used in desktop grids. However, the existing desktop grid systems have mainly used static task scheduling such as first-come first-served (FCFS) scheduling [21,22] and eager scheduling [11,23]. Such systems cannot gua...

show abstract

Dynamic scheduling for heterogeneous Desktop Grids

Cited by 11 publications

References 15 publications

Whare-map

Whare-map

Bag-of-Tasks applications scheduling on volunteer desktop grids with adaptive information dissemination

Task scheduling scheme based on resource clustering in desktop grids

Contact Info

Product

Resources

About