Introduction - Cloud Data Management - page 7

on performance characteristics from choosing to operate with weak consistency
mechanisms. The overall methodology of experiments, for measuring consistency
from a customer's view, is also another contribution.
Chapter 6 describes a solution to replication evaluation on virtualized database
servers. In addition to the two widespread approaches, namely NoSQL database as
a service and relational database as a service, virtualized database servers is the
third approach for deploying data-intensive applications in cloud platforms. It takes
advantages of virtualization technologies by taking an existing application designed
for a conventional data center and then porting it to virtual machines in the public
cloud. Such migration process usually requires minimal changes in the architecture
or the code of the deployed application. In this topic, the limits to scaling for an
application that itself manages database replicas in virtualized database servers in
the cloud is explored. A few important limits are characterized in the load on the
master copy, the workload imposed on each slave copy when processing updates
from the master, and also from the increasing staleness of replicas.
Chapter 7 introduces a SLA-driven framework for managing database replica-
tion. Cloud-hosted database systems, such as virtualized database servers, powering
cloud-hosted applications form a critical component in the software stack of these
applications. However, the specifications of existing SLA for cloud services are
not designed to flexibly handle even relatively straightforward performance and
technical requirements of customer applications. Motivated by this, in this topic, a
novel adaptive approach for SLA-based management of virtualized database servers
from the customer perspective is presented. The framework is database platform-
agnostic, supports virtualized database servers, and requires zero source code
changes of the cloud-hosted software applications. It facilitates dynamic provision-
ing of the database tier in software stacks based on application-defined policies for
satisfying their own SLA performance requirements, avoiding the cost of any SLA
violation and controlling the monetary cost of the allocated computing resources.
Therefore, the framework is able to keep several virtualized database replica
servers in different data centers to support the different availability, scalability and
performance improvement goals. The experimental results confirm the effectiveness
of the SLA-based framework in providing the customer applications with the
required flexibility for achieving their SLA requirements.
Chapter 8 presents a genetic-algorithm-based service composition approach
cloud computing. In particular, a coherent way to calculate the QoS values of
services in cloud computing is presented. In addition, comparisons between the
proposed approach and other approaches show the effectiveness and efficiency
of the proposed approach. Chapter 9 provides a comprehensive overview for
modern approaches and mechanisms of large scale data processing mechanisms and
systems. Chapter 10 concludes the contents of this topics and sheds the lights on
a set of research challenges that have been introduced by the new wave of cloud-
hosted data storage and big data processing systems.