Cloud infrastructure resource management

云基础设施资源管理

• 批准号: 486584-2015
• 负责人: Agarwal, Anjali
• 金额: $ 2.19万
• 依托单位: Concordia University
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2016
• 资助国家: 加拿大
• 起止时间: 2016-01-01 至 2017-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=600312
• 关键词: Cloud; infrastructure; resource; management

Cloud computing is a common name in IT industry. However increase in Cloud computing usage also comes with increased complexity. Efficient resource management for best utilization with minimal cost is a crucial task. The scope of this work is under IaaS service module, which handles infrastructure resources (Virtual Machines, Servers, Storage, and Network) allocation. Resource management offers the following benefits: elasticity/scalability, QoS, optimal utility, reduced overhead, improved throughput, reduced latency, cost effectiveness and simplified interface. VM migration is one of the most popular way to manage resources, and Live VM migration is the most used technique to reload or rearrange the resources in datacenter to keep the delivered services available. Statistics shows 70% of datacenters resources is not utilized. Therefore, reduction in number of servers and hardware will not only reduce its capital costs, but will also reduce its running and maintenance cost. In addition, Live VM migration will facilitate load balancing and energy management as part of resource management. VM migration consists of optimally selecting VM to be migrated, which physical server must be chosen to move this VM, when to start migration process, and duration of migration process. The goals for VM Selection techniques are to reduce power consumption, increase load balancing, and increase fault tolerance, which eventually will increase the cloud services availability and throughput/efficiency. The increased efficiency reduces the overall operation costs, pollution (green datacenters), and hardware maintenance. Migration Process itself targets moving the VM in minimum time to avoid any interruption of services. In this research proposal, in collaboration with Cistech Ltd, we consider investigating solutions to decrease the migration process time and optimally selecting VMs for live migration. Our project goal is to develop smart algorithms that will evaluate the available running resources for load balancing, server consolidation, network bandwidth, VMs shrinking or expanding and fault tolerance. Our project will drive the energy saving and the cost saving in running the IT datacenters.

云计算是IT行业的一个通用名称。然而，云计算使用的增加也带来了复杂性的增加。以最小的成本实现最佳利用的有效资源管理是一项至关重要的任务。这项工作的范围在IaaS服务模块下，该模块处理基础设施资源（虚拟机、服务器、存储和网络）分配。资源管理提供了以下好处：弹性/可扩展性、QoS、最佳效用、降低开销、提高吞吐量、降低延迟、成本效益和简化接口。虚拟机迁移是管理资源的最流行的方法之一，而实时虚拟机迁移是最常用的技术，用于重新加载或重新排列数据中心中的资源，以保持交付的服务可用。统计显示，70%的企业资源没有得到利用。因此，减少服务器和硬件的数量不仅可以降低其资本成本，还可以降低其运行和维护成本。此外，动态VM迁移将促进负载平衡和能源管理，作为资源管理的一部分。VM迁移包括最佳选择要迁移的VM、必须选择哪个物理服务器来移动此VM、何时开始迁移过程以及迁移过程的持续时间。VM选择技术的目标是降低功耗，增加负载平衡，并提高容错能力，最终将提高云服务的可用性和吞吐量/效率。提高的效率降低了总体运行成本、污染（绿色排放）和硬件维护。迁移过程本身的目标是在最短的时间内移动VM，以避免任何服务中断。在本研究提案中，我们与CiscoTech Ltd合作，考虑研究解决方案以减少迁移过程时间，并为实时迁移选择最佳虚拟机。我们的项目目标是开发智能算法，用于评估负载平衡、服务器整合、网络带宽、虚拟机收缩或扩展以及容错的可用运行资源。我们的项目将推动节能和成本节约运行的IT中心。