A Survey on Dynamic Optimization of Multi attributes Resource Allocation in Self- Organizing Clouds
|Richu Jimmy 1, S. Vanitha 2, Daniel .D 3
|Related article at Pubmed, Scholar Google|
Virtual Machine is a software implementation of a computing environment in which operating system can be installed & run. By combining the concept of volunteer computing with cloud architectures, we can build self organizing clouds (SOC). It gathers unused resources over internet. Here each participant can act as both resource provider and requester. Fully distributed, Virtual Machine multiplexing resource allocation scheme is used to manage decentralized resources. Proportional share model aims at providing maximized resource utilization and optimal execution efficiency. Pointer-Gossiping CAN (PG-CAN) is used to locate qualified nodes with minimum contention. Existing solutions produces heavy weight query messages per tasks. These messages will utilize large amount of resources for computation. The above algorithms along with PG-CAN set up are used in order to produce light weight queries per task in our SOC. This set up can produce an improvement in system output by 15-60 percent than peer to peer model which was used earlier.