Migrating e-Science Applications to the Cloud: Methodology and Evaluation - Cloud Computing with e-Science Applications - page 106

HTML (hypertext markup language), JavaScript, JSP, and CSS. The busi-
ness logic layer is implemented in Java. For the object-relational mapping,
we use Java Data Objects version 3.1 and its implementation DataNucleus
version 3.0 (http://www.datanucleus.org). For online hosting of the tool we
use Google Cloud SQL as the data layer and run the whole application in
Google's App Engine. A stand-alone, offline version of the tool also exists,
allowing the user to run the tool locally. In this case, MySQL 5.5 is used for
the data layer and Apache Tomcat version 7 as the servlet container. Further
information is available on the website of the Cloud Data Migration Tool
(http://www.cloud-data-migration.com).
5.5 Evaluation
In this section, we evaluate both the methodology introduced in Section 5.4.2
and the Cloud Data Migration Tool supporting this methodology presented
in the previous section. For this purpose, we used the motivating scenario
discussed in Section 5.2 as a case study involving the migration of the
database layer of the SimTech SWfMS to the cloud.
As our investigation of the literature did not result in a method that spe-
cifically aims at the evaluation of migration methodologies, we focused our
analysis on related evaluation methods and standards for software processes
and software quality. Al-Qutaish and Berander et al. provided an over-
view of available software quality models and standards (Al-Qutaish, 2010;
Berander et al., 2005). Based on their findings, we selected the International
Organization for Standardization/International Electrotechnical Com-
mission (ISO/IEC) 9126 standard provided for the evaluation of the Cloud
Data Migration Tool, as its quality attribute model includes the metrics we
considered most relevant, such as understandability and operability (Jung,
Kim, and Chung, 2004). For the evaluation of software processes, there are
multiple guidelines (e.g., Shull, Carver, and Travassos, 2001; Sommerville,
1996) and standardized best practices, such as Capability Maturity Model
Integration (CMMI) (CMMI Product Team, 2010) and the Continual Service
Improvement (CSI) module of the IT Infrastructure Library (ITIL) (Case and
Spalding, 2011). We based our evaluation of the migration methodology on
the ITIL CSI process but adapted it to consider the technical aspects of the
methodology by considering appropriate metrics for software processes pro-
vided by Daniel (2004). A simplified representation of the resulting process
is shown in Figure 5.5.
In the first step, a strategy for the realization of the process was deter-
mined. In this case, our strategy was to use the Cloud Data Migration Tool
discussed in the previous section in conjunction with a specific migration