Phd Projects | IEEE Project | IEEE Projects 2020-19 in Trichy & Chennai

IEEE Projects Trichy, Best IEEE Project Centre Chennai, Final Year Projects in Trichy - We Provide IEEE projects 2018 - 2019 , IEEE 2018 Java Projects for M.E/M.Tech, IEEE 2018 Dot net Projects for B.E/B.Tech, IEEE 2018 Power electronics Projects Engineering & Diploma Students, Matlab, Embedded, NS2 Projects
You are here: Home / ieee projects 2013 / Heuristic load balancing Method for improving the task scheduling

Heuristic load balancing Method for improving the task scheduling

by

SOC is different from the traditional Grid model (including P2P desktop Grid) in the resource consumption manner. Grids generally assume exclusive resource usage to ensure user QoS. The problem of job scheduling in Grids is usually categorized as a multiprocessor scheduling (MPS) problem (a kind of combinatorial optimization problem), which has been proved to be NP-complete. Accordingly, many approximation algorithms as well as (meta) heuristics applied to various versions of the MPS problem in the Grid environment have been studied. CLOUD computing has emerged as a compelling paradigm for deploying distributed services. Resource allocation problem in cloud systems emphasizes how to harness the multiattribute resources by multiplexing operating systems. With virtual machine (VM) technology, we are able to multiplex several operating systems on the same hardware and allow task execution over its VM substrates without performance interference. Fine-grained resource sharing can be achieved as each VM substrate can be configured with proper shares of resources (such as CPU, memory, storage, network bandwidth) dynamically. Proposed a heuristic load balancing method for improving the task scheduling throughput on desktop Grids over CAN overlay. The problem to be a bins-and-balls model with herds phenomenon and tried to get the approximately optimal performance using a stochastic algorithm atop a DHT overlay. Studied a user-centric utility function of task turnaround time to improve the system performance based on simulation. Compared with these existing works, we devise an autonomous VM-multiplexing resource consumption model, namely SOC, which allows each task to dynamically make full use of the resource slices isolated by VM technology. In SOC, every task is allowed to share the multiple types of resources on the same node, so the resources can be utilized more abundantly. For example, a CPU-bound task and an IO-bound task could run at the same physical node at the same time by leveraging VM resource isolation technology