Adaptive Execution of Scientific Workflow Applications on Clouds - Cloud Computing with e-Science Applications - page 79

tasks so the cost-benefit analysis of utilization of the cloud can be
undertaken and the utilization of cloud computing for workflows
can be justified.
These requirements were addressed while we developed an adaptive
system for execution of a workflow for agent-based simulation in hybrid
clouds. The application is detailed next.
4.4 Case Study
A city is sustainable only if it can accommodate economic and population
growth while ensuring the well-being of its people and environment [9]. There-
fore, reaching sustainability becomes harder when the growth of a population
is high or when the growth occurs in areas of high density, such as Singapore.
Singapore's land area has increased from 581 km 2 in the 1960s to 716 km 2
in 2012; its population in the same period has grown from 1.6 million to
5.3 million [10]. To maintain reasonably good economic growth, the Singapore
government has projected a need for the population to reach 6.9 million by
2030. However, the land area is only slated to grow to about 800 km 2 in the
same period. The disparity in the growth rate of population versus land area
means that there is increasing strain on space and the service infrastruc-
ture. It is crucial for the planning agencies to adopt a scientific approach to
understanding the urban fabric and how it can adapt to social, economic,
and environmental changes.
One key aspect to improve the quality of living of city inhabitants is public
transport. There is a need for efficient transport covering the biggest exten-
sion of the city as possible and running with enough frequency so people are
motivated to use it rather than using cars. In this sense, Singapore's public
transport network is ranked among the best in the world. Its Mass Rapid
Transit (MRT) train network comprises 102 stations distributed over four
main lines, with a total of almost 150 km of rail lines. It currently serves
around 2.5 million commuters per day, which represents more than 75% of
the total public transport users [11].
The number of commuters and the high frequency of trains (running in
intervals as short as 90 seconds) make it a complex system. Furthermore,
even a minimal disruption in the operation of one train can cascade over
several lines, affecting hundreds of thousands of commuters.
This complex and sensitive system will be subject to even further pressure
as the population increases. Therefore, tools are needed to help planners
evaluate the effects that disruptions would have over the whole system.