@ Boston University
Trustworthiness is a nebulous concept, often used to capture various considerations, spanning dimensions of performance, safety, reliability, scalability, predictability, security, and privacy, among others. While cloud virtualization promotes many of these considerations, others remain elusive. This state of affairs is primarily due to the fact that we do not possess (yet) a good handle on how to define and verify the trustworthiness dimensions of cloud computing, how to design mechanisms that deliver desirable levels of trustworthiness along each of these dimensions as well as enable tradeoffs across these dimensions, and how to expose these tradeoffs to cloud customers and system integrators in a ways that are both practical and usable.
To be trustworthy, cloud services must also be delivered at a verifiably “fair market value”. This requires that cloud resources be viewed as marketplace commodities that various independent and rational cloud users must compete for. Such a mechanism is essential for trusting the (economic) value proposition of deploying services in the cloud. Thus, resource reservation and performance isolation mechanisms afforded by virtualization as well as mechanisms supporting the aforementioned security and integrity considerations must be complemented with mechanisms that ensure that the cost bourn by an individual user (application) is justifiable by “market” conditions based on supply (availability of cloud resources of various capabilities) and demand.