Deadlock prevention for resource allocation in model nVM-out-of-1PM

Ha Huy Cuong Nguyen, Van Son Le, Thanh Thuy Nguyen · 2016

The cloud computing paradigm shift brings about disruptive changes to the traditional business models in the infocomm sector, especially for existing software licensing models and server purchase or lease models. Vietnam Posts and Telecommunications Group (VNPT) have ramped up their cloud computing efforts as a strategic response to protect their current market share. This paradigm shift also has major separability on Vietnam's infocomm industry, which is heavily dependent on these traditional VNPT. Since 2014, Viet Nam has started offering services in domestic and international network in the country. The Global Data Services Joint Stock Company (GDS) is a joint venture between NTT Communications (NTT Com) and Vietnam Posts Telecommunications (VNPT). Although, a number of studies were published by third parties on how Vietnam has fared in its cloud journey. While such studies use different definitions and yardsticks, their results nonetheless are instructive in helping us to understand the progress Vietnam has made and to highlight areas for improvement. We have a few studies resources allocation such infrastructure as a services. With the aim to share research solutions and seek opportunities for local businesses who want to invest in infrastructure cloud from the Open Source duty. In this paper, we proposes an algorithm deadlock prevention the model n VM - out - of 1 PM to separate and manage resources. The algorithms for allocating multiple resources to competing services running in virtual machines on a heterogeneous distributed platforms. This paper presents an algorithm prevention deadlock used to reschedule the policies of resource supply for resource allocation. In the current scenario, deadlock prevention algorithm using method global state prevention based. We have implemented and performed our algorithm proposed by using CloudSim simulator.

Read the paper · More papers on PaperTik