21.01.2014 Views

Xiao Liu PhD Thesis.pdf - Faculty of Information and Communication ...

Xiao Liu PhD Thesis.pdf - Faculty of Information and Communication ...

Xiao Liu PhD Thesis.pdf - Faculty of Information and Communication ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Abstract<br />

Cloud computing is a latest market-oriented computing paradigm which can provide<br />

virtually unlimited scalable high performance computing resources. As a type <strong>of</strong><br />

high-level middleware services for cloud computing, cloud workflow systems are a<br />

research frontier for both cloud computing <strong>and</strong> workflow technologies. Cloud<br />

workflows <strong>of</strong>ten underlie many large scale data/computation intensive e-science<br />

applications such as earthquake modelling, weather forecast <strong>and</strong> Astrophysics. At<br />

build-time modelling stage, these sophisticated processes are modelled or redesigned<br />

as cloud workflow specifications which normally contain the functional<br />

requirements for a large number <strong>of</strong> workflow activities <strong>and</strong> their non-functional<br />

requirements such as Quality <strong>of</strong> Service (QoS) constraints. At runtime execution<br />

stage, cloud workflow instances are executed by employing the supercomputing <strong>and</strong><br />

data sharing ability <strong>of</strong> the underlying cloud computing infrastructures. In this thesis,<br />

we focus on scientific cloud workflow systems.<br />

In the real world, many scientific applications need to be time constrained, i.e.<br />

they are required to be completed by satisfying a set <strong>of</strong> temporal constraints such as<br />

local temporal constraints (milestones) <strong>and</strong> global temporal constraints (deadlines).<br />

Meanwhile, task execution time (or activity duration), as one <strong>of</strong> the basic<br />

measurements for system performance, <strong>of</strong>ten needs to be monitored <strong>and</strong> controlled<br />

by specific system management mechanisms. Therefore, how to ensure satisfactory<br />

temporal correctness (high temporal QoS), i.e. how to guarantee on-time completion<br />

<strong>of</strong> most, if not all, workflow applications, is a critical issue for enhancing the overall<br />

performance <strong>and</strong> usability <strong>of</strong> scientific cloud workflow systems.<br />

At present, workflow temporal verification is a key research area which focuses<br />

on time-constrained large-scale complex workflow applications in distributed high<br />

performance computing environments. However, existing studies mainly emphasise<br />

IV

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!