J. Srinivasan1 · C. Suresh Gnana Dhas2 Received: 20 February 2020 / Accepted: 24 April 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Cloud computing (CC) is an attractive emerging technology due to offering services based on-demand by the process of virtualization. Since CC platform offers services based on-demand it has been widely used in the field of various emerging IT infrastructure. In cloud platform each application is run in individual virtual machine (VM) for execution of services within the host. Since cloud platform operates on on-demand service it need to cope with multiple application in single time hence it is necessary to adopt an effective approach for balancing memory utilization in cloud network. For effective utilization of available memory existing approaches uses probability distribution method for allocating resources in cloud platform but still there exists a lack of utilization of available memory in cloud platform. This paper aims to develop an effective approach for dynamic memory allocation in VM in cloud platform. For memory allocation among VM in cloud platform proposed approach uses cloud vertical elasticity manager (CVEM), memory reporter (MR), memory over subscription granter (MOG). The MOG uses a scheduler to allocate the memory in a dynamic way inside a host. Finally, we adopt host elasticity rule to balance the available memory to allocate dynamically the memory inside an available host in cloud. Keywords Cloud vertical elasticity manager · Memory reporter · Memory over subscription granter · Dynamic memory allocation
In recent years, cloud computing (CC) is emerging information technology (IT) service delivery model offers essential characteristics like elasticity, resource on demand, broad network access, resource pooling, and measured service (Beltran et al. 2016). These characteristics makes the CC services as an emerging technology. Various technologies evolved in CC services includes distributed computing, virtualization, service-oriented design, utility computing, storage, networking, etc. (Baranwal and Vidyarthi et al. 2015). Out of which, the resource allocation (RA) algorithms is considered as an important target in recent era (Yuvaraj and Suresh Ghana Dhas 2018; Yuvaraj et al. 2013, 2018, 2019; Sivaram et al. 2019). The target is regarded as a group of services that understands a service in the form of a separable job onto VM groups. The resource usage is optimized to offer better quality of service (QoS) and it further limits the migrations. Various solutions in this perspective is developed for the purpose of RAs and this includes bin packing algorithms. Even though there is no clear agreement within the community on that aspects of the matter area unit, the necessity on an algorithmic model requires such consideration (Beaumont et al. 2016). The streamlined edges of central desk-based operation, commonplace protected installation, data security and simpler clients control on virtual desktop-clouds (VDCs) (Yan 2011) have now become increasingly obvious to the business community virtualisation by Desktop-as-a-Service (DaaS) ofers. Above all, in virtual desktops (VDs), the expected client QoE is prone to network loss and cannot withstand disruptive patterns of cross traffic (Fiedler et al. 2010; Calyam et al. 2014).