An Economic-based Resource Management and Scheduling for Grid Computing Applications

20 An Economic-based Resource Management and Scheduling for Grid Computing Applications

G. Murugesan1, Dr.C.Chellappan2

1 Research Scholar, Department of Computer Science and Engineering,
Anna University, Chennai-600 025, Tamil Nadu, India

2 Professor, Department of Computer Science and Engineering,
Anna University, Chennai-600 025, Tamil Nadu, India

Abstract Resource management and scheduling plays a crucial role in achieving high utilization of resources in grid computing environments. Due to heterogeneity of resources, scheduling an application is significantly complicated and challenging task in grid system. Most of the researches in this area are mainly focused on to improve the performance of the grid system. There were some allocation model has been proposed based on divisible load theory with different type of workloads and a single originating processor. In this paper we introduce a new resource allocation model with multiple load originating processors as an economic model. Solutions for an optimal allocation of fraction of loads to nodes obtained to minimize the cost of the grid users via linear programming approach. It is found that the resource allocation model can efficiently and effectively allocate workloads to proper resources. Experimental results showed that the proposed model obtained the better solution in terms of cost and time.
.

Keywords: Grid Scheduling, Resource management, Workload distribution, Economic model, Cost Optimization

1. Introduction One of the most complicated task in Grid computing is the allocation of resources for a process; ie., mapping of jobs to various resources. This may be a NP-Complete (Non-deterministic Polynomial time) problem. For example, mapping of 50 jobs into 10 resources produces 1050 possible mappings. This is because every job can be mapped to any of the resources. In our case the allocation is in terms of co-allocation which means that the job is executed on a number of resources instead of single resource. Here resource means processors which are involved in the scheduling process. We used resources and processors simultaneously. The other complexity of resource allocation is the lack of accurate information about the status of the resources. Load balancing and scheduling play a crucial role in achieving utilization of resources in grid environments [20].

Much of the work was done on finding an optimal allocation of resources in Grid computing environments. The scheduling schemes are divided into two main categories; conventional and economical. The conventional strategies consider the overall performance of the system as a metric for determining the system quality. It does not take the cost as factor for scheduling jobs on resources and treat all resources as the same at all. Some examples are SmartNet, AppleS Project, Condor-G, NetSolve etc. In economic strategy, cost is considered as essential factor for scheduling jobs. The user is charged based on the utility of the resources in the Grid system. Some of the works consider the economic strategies which deals with the price of resources when it needs to allocate jobs to resources and that price usually reflects the value of the resource to the user.

Task scheduling is an integrated part of parallel and distributed computing. The Grid scheduling is responsible for resource discovery, resources selection, job assignment and aggregation of group of resources over a decentralized heterogeneous system; the resources belong to multiple administrative domains. The resources are requested by a Grid application, which use to computing, data and network resources etc. However, Scheduling an applications of a Grid system is absolutely more complex than scheduling an applications of a single computer. Because to get the resources information of single computer and scheduling is easy, such as CPU frequency, number of CPU’s in a machine, memory size, memory configuration and network bandwidth and other resources connected in the system. But Grid environment is dynamic resources sharing and distributing. Then an application is hard to get resources information, such as CPU load, available memory, available network capacity etc. And Grid environment also hard to classify jobs characteristic, that run in Grid. There are basically two approaches to solve this problems, the first is based on jobs characteristic and second is based on a distributed resources discovery and allocation system. It should optimize the allocation of a job allowing the execution on the optimization of resources. The scheduling in Grid environment has to satisfy a number of constraints on different problems.

The existing scheduler used in TeraGrid and other notable compu

…(Full text truncated)…