Irrespective of the deployment model, in general any organization opting for cloud must consider the following implications:
Network Dependency – Whether you choose, on-site or off-shore, a reliable and secure network is highly desirable for good performance.
Subscribers still need IT skills – You can’t just offer a pink-slip to all your IT resources. To manage various user devices that access cloud, resources with traditional IT skills are required, though in lesser number. Additionally, your existing resources may need to update themselves with new skills for working in cloud.
Risk from multi-tenancy – On-site private cloud mitigates this security risk by restricting the number of possible attackers as all the clients are typically the members of one subscriber organization. In a public cloud scenario, a single machine may be shared by the workloads of any combination of subscribers. This indeed raises the security risk as the number of potential attackers increases with number of subscribers. Therefore we can safely conclude that risk due to multi-tenancy increases in an order which can be stated as Private, Community, Hybrid, Public cloud.
Data import/export and performance limitations – Generally the on-demand bulk data import/export is limited by the cloud’s network capacity. In the on-site
private cloud scenario, however, these limits may be adjusted, although not eliminated, by provisioning high-performance and/or high-reliability networking within the subscriber's infrastructure.
private cloud scenario, however, these limits may be adjusted, although not eliminated, by provisioning high-performance and/or high-reliability networking within the subscriber's infrastructure.
Workloads Locations – Workloads refers to managing hardware resources efficiently. Generally, cloud migrates workloads between machines without any inconvenience to the clients, i.e., it’s hidden from the client. Generally, the cloud vendors take care of this but you must explicitly check with your vendor if it manages the resources efficiently.
The implications described here are general in nature. Before making any decision in favor of a specific deployment model, study the detailed implications of that particular deployment model. For details, please check the reference section.
