Evaluation Criteria to Run Scientific Applications in the Cloud - Cloud Computing with e-Science Applications - page 7

TABLE 1.1
Comparison of Implementation Models

Advantage Drawback

Private Privacy Scalability

Community Shared cost Scalability

Public Scalability Privacy

Hybrid Scalability Interconnection

1.3.5 Summary of the Implementation Models

Summarizing the characteristics presented in this section, we can conclude
that all deployment models can be used to create high-performance comput-
ing (HPC) environments in the cloud. The appropriate model depends on the
needs of the user and the user's available funds. All the models have advan-
tages and disadvantages, and it is clear that there is no ideal model for all the
usage scenarios. Table 1.1 summarizes the main advantage and disadvantage
of each cloud implementation model.

1.4 Considerations about Public Providers
The private and community models are well known by users due to their
similarity to clusters and grids. The hybrid and public models are really
new paradigms of computing. As the hybrid model is a combination of local
machines and a public provider, we can conclude that the new paradigm is
the public cloud. In the rest of this chapter, we perform an analysis of the
public cloud model.
When choosing a public cloud provider, the user needs to consider relevant
aspects of his service. Some of these concerns are explained here. However, the
user needs to perform an analysis of the necessary service level for his service.

1.4.1 Data Confidentiality

Data confidentiality is one of the main concerns regarding public cloud
providers. In addition, relevant aspects about data manipulation need to
be considered:

• Segregation : The provider needs to guarantee data segregation
between clients because most of them use shared resources. It is
necessary to ensure that the user's data can only be accessed by
authorized users.