Reference no: EM133182680

Questions:

From the first angle, how can the cloud help in answering difficult research questions? Can data-intensive applications provide knowledge and answers that could open new frontiers of our understanding? While this is the main driver for research and development of grid computing architectures, it is still unclear how to optimally operate a cloud system in scientific domains, such as physics and engineering, for example. Also, how can large-scale computation be achieved in a reliable and efficient manner? The body of work devoted to high-performance computing strives to continuously improve efficient and effective computational and parallel processing models.

Second, what are the ways to improve cloud services and architecture? Can cloud computing serve a larger number of users in a consistently transparent yet reliable manner?

Most recent work has focused on improved service provisioning, tackling problems related to parallelization, scalability, efficiency, and large-scale processing, along with monitoring and service control of data-intensive applications. As noted by Barker et al., there are some important opportunities for research in cloud computing that require further exploration. These include user-driven research (how to develop environments that support budget-limited computation based on a set of user-driven requirements), new programming models (what are, if any, the alternatives to MapReduce?), PaaS environments, and improved tools to support elasticity and large-scale debugging.

Finally, how can we improve cloud adopters' confidence and limit potential risks from using cloud services? Some recent statistics have shown users' reluctance in adopting clouds due to a lack of confidence in the security guarantees offered by cloud providers, and in particular, poor transparency. Specific issues reported by users related to lack of confidentiality, poor integrity guarantees, and potentially limited availability.