Database Reference
In-Depth Information
minimum replication that needs to be created can be determined. As a direct consequence,
the minimum replication can also act as a benchmark, which can be used for evaluating the
cost-effectiveness of various replication-based data storage approaches.
3. Cost-effective data reliability assurance
In order to maintain the Cloud data with the minimum replication level, a mechanism that is
able to create Cloud data based on the minimum replication calculation approach as well as
maintain the created replicas in the Cloud needs to be designed. For effective Cloud data re-
liability management, this mechanism needs to be able to maintain the big data in the Cloud
with a wide variety of data reliability assurance so that all different levels of data reliability
requirements can be met. In addition, as a very important aspect, the overhead of such a
mechanism also needs to be taken into account.
4. Cost-effective data transfer
When replicas of the Cloud data need to be created or are lost, we need to provide effective
data transfer process that could maintain the replication level of the data in a cost-effective
fashion. In the data creation and recovery stages of the Cloud data life cycle, data transfer
activity plays the major role, which transfers the data to the appropriate storage devices.
Therefore, optimizing the data transfer in Cloud network could be a good solution for cost-
effectiveness. By optimizing data transfer, the cost incurred by data creation or recovery can
be reduced.
1.4
Book overview
This topic systematically investigates the challenging issue of providing cost-effective
data storage with data reliability assurance, which includes solid theorems and practi-
cal algorithms and finally forms a comprehensive solution to deal with the issue. The
topic structure is depicted in Figure 1.2 .
In Chapter 2 , we introduce existing works in literature related to our research. To
facilitate our research, literature in three major fields are reviewed. First, from the
hardware aspect, to investigate the reliability pattern of storage devices in the Cloud,
literature on hardware reliability theories are reviewed. Second, from the software
aspect, to investigate data reliability models, and data redundancy maintenance ap-
proaches in the Cloud, literature on data reliability modeling, data reliability assurance
approaches in distributed data storage systems are reviewed. Third, to investigate data
recovery approaches in the Cloud, literature on data recovery and data transfer ap-
proaches in distributed systems are reviewed.
In Chapter 3 , we present the motivating example of this topic and analyze our re-
search problem. We first introduce the motivating example of our research, which is a
real-world scientific application for pulsar searching survey of typical data-intensive
characteristics. Based on the motivating example, we analyze the research problem
and identify details of our research issues.
In Chapter 4 , we present our data reliability model for Cloud data storage. Based
on the details of our research issues identified in Chapter 3 , first we further determine
several properties for our data reliability model, and then our novel generic replica-
tion-based data reliability model is presented in detail.
In Chapter 5 , we present the minimum replication calculation approach. Based
 
Search WWH ::




Custom Search