Article Preview
TopIntroduction
Cloud Computing is an on-demand, scalable, elastic, transparent service created to optimize and virtualize resources with a gradual evolution of distributed, parallel, and grid computing services. It enables access to high-end computations without considering infrastructure, platform, or software. It allows many micro and mini enterprises to innovate and own services required at a nominal cost without thinking about the maintenance and deployment costs. In other words, Cloud Computing is location and device-independent, enabling the services and data to be available at all times and places.
Cloud Computing has numerous benefits for both customers and hosts. Characteristics of Cloud Computing are more complex and futuristic than grid computing or distributed computing models (Singh, Agarwal, & Mishra, 2017). Cloud characteristics discussed by various authors include Resource Pooling, On-demand Self-Service, Availability, Economical, Pay-as-you-go, Easy Maintenance, Service negotiations, Large Network Access, Automatic Systems, Security (Apte, Hu, Schwan, & Ghosh, 2010; Böhm, Leimeister, Riedl, & Krcmar, 2010; Callegati & Cerroni, 2013; Carrasco, Durán, & Pimentel, 2020; Chen, Chiew, Ye, Zhu, & Chen, 2013; Demchenko, Turkmen, Slawik, & De Laat, 2017; Fan, et al., 2018).
Hardware and virtualization architectures help to achieve portability and interoperability of the application and its dependencies. Portability refers to moving an application or data between systems yet suitable to adapt the application. In comparison, an application’s ability to run on diverse systems or components is interoperability. A different layer of the Cloud serves the application, service, and data. As discussed by (Council, 2017), the computing portability and interoperability categories include Data, Application, Platform Portability and Application, Service, Platform, Management, Publication and Acquisition Interoperability (Mell & Grance, 2011).
Virtualization allows cloud users to exploit multi-tenant resources (Wallis, 2008). Multi-tenancy is very useful, especially in public clouds, as these cloud providers often have different or customized hypervisors. The customized hypervisors contribute to the heterogeneity of the Cloud. However, the data and application get locked once they get deployed on Cloud to a specific vendor. This problem is called vendor lock-in. Vendor Lock-In outlines the restricted use of technology, a service solution developed by the vendor for a particular service (Opara-Martins, Sahandi, & Tian, 2015). Vendor Lock-In identifies itself as one of the most critical and crucial issues in cloud computing. It poses a risk of migration and restrains the business to switch from the existing system to Cloud or across Clouds. The National Institute of Standards and Technology (NIST) clearly defines and distinguishes service interoperability, system portability, and data portability (Bouzerzour, Ghazouani, & Slimani, 2020).
Migration is a technique of transferring the application or its data from one facility to another. Data can be of any form like audio, videos, images, documents, texts, datasets, application files, and many more. Cloud providers like Google use Kernel-based Virtual Machine (KVM), providers like Amazon Web Services (AWS) or Rackspace use Xen. Microsoft uses Hyper-V for its Windows Azure services. Even if different cloud providers use the same hypervisor for virtualization, the Migration of data and applications is highly challenging, especially when Cloud users are live (Mansour & Bouchachia, 2017). During this time, the users might experience the unavailability of resources, which is significantly proportional to performance hindrance. The Virtual Machine (VM) stops during the migration process’s total duration, which causes service interruption to the user. Interoperability and portability are the two solutions for solving the above-said problems (Council, 2017).