Faculty of Engineering and Technology
Coursework Title: Data centre design and analysis via simulation
Module Name: Cloud Computing
Module Code: 6046COMP
Level: FHEQ6
Credit Rating: 24
Weighting: 100%
Maximum mark available: 100
Lecturer: Dr Gabor Kecskemeti
Contact: If you have any issues with this coursework you may contact your lecturer.
Contact details are:
Email:
Room: BS/701
Hand-out Date: 22nd Jan 2018
Hand-in Date: 17th Apr 2018
Hand-in Method: Canvas
Feedback Date: 8 May 2018
Feedback Method: Email
Programmes:
Introduction
This coursework is a research based assessment on your understanding of data centre design and
virtualisation in the context of Cloud Computing. The aims of the coursework are:
• To demonstrate your ability to specify and design a Cloud Computing capable data centre.
• To show an understanding of key Cloud Computing technologies
Learning Outcome to be assessed
1. Describe the hardware and software concepts and architecture of Cloud Computing.
2. Contrast the key technical and commercial issues concerning Cloud Computing versus
traditional software models.
3. Recognise the importance of virtualisation technology in support of Cloud Computing.
4. Specify and design Cloud Computing capable data centres.
Detail of the task
Traditional data centres are buildings where multiple servers and associated communications
resources are housed together for ease of maintenance and because of common environmental
and physical security needs. In this sense Cloud Computing is viewed as data centre computing.
Traditional data centres, however, usually comprise of a large number of reasonably small
applications hosted on dedicated hardware, partitioned from other systems in the same data
centre. There is a heterogeneous mix of hardware, software and associated maintenance needs
with little communication between separate systems. In contrast the cloud computing offerings from
companies such as Microsoft, Google, Amazon and Yahoo are owned and operated by a single
company, employ mostly homogeneous hardware and software platforms and are administered
through a common systems’ management layer.
Based on the research papers provided (available from the module’s Canvas site) and your own
research explain the major factors associated with Cloud Computing data centre design and
management. Comment particularly on data centre architectures, warehouse scale data centres
and modular data centres; reporting on their contribution to data centre scalability, performance
and dependability. You should make reference to virtualisation, where appropriate, commenting on
its use and operation in a cloud computing context. After completing the research, use the insights
you have gained to propose a data centre design. Consider both the architectural design for
computing clouds, in the context of the physical requirements of the data centre, such as power,
cooling, reliability, rack space and location, and with emphasis on distributed computing including
design goals and enabling technology.
Finally, for higher marks, present your data centre design in a cloud simulator (preferably
DISSECT-CF – which is available at https://github.com/kecskemeti/dissect-cf). In this final part of
the coursework you are expected to describe the datacentre of yours in terms of source code. This
source code should construct your previously designed data centre with the components of your
chosen simulator. As part of this last task, to show that your work is really behaving as a cloud
data-centre, analyse your simulated data-centre’s behaviour by executing traditional and cloud
workloads on it and provide an overall discussion on the observed behaviour in your handed in
word document. In this final part of your design document you are expected to make sure that your
research is thoroughly reflected in the actual simulation as well. The better the connections
between the documentation and the source code is the higher the marks you could expect.
What you should hand in
A word processed document (preferably in pdf form) and simulation related source code in a single
zip file through the assignment handler link on Canvas.
Your document should be around 10-15 pages in length. Overlong reports will not gain high marks:
Marks will be awarded for explaining the concepts in a concise manner. The final mark will be
determined by the quality of the work.
Assessment Criteria weighting for
each problem part
1. Data Centre Design oriented research
a. Data Centre Architecture Discussion in contrast with traditional
data-centres
b. Warehouse scale and Modular Data Centres Distinction
c. Scalability, Performance and Dependabilty Discussion
d. Virtualisation
1. 40
a. 10
b. 10
c. 10
d. 10
2. Data Centre Design
a. Distributed computing architectural design and enabling
technologies
b. Physical requirements
c. Simulated model of the designed data-centre
d. Behaviour analysis of the data-centre with various workloads
2. 60
a. 10
b. 10
c. 20
d. 20
Recommended reading
1. Bari, M.F., Boutaba, R., Esteves, R., Granville, L.Z., Podlesny, M., Rabbani, M.G., Zhang,
Q. and Zhani, M.F., 2013. Data center network virtualization: A survey. Communications
Surveys & Tutorials, IEEE, 15(2), pp.909-928.
2. Al-Fares, M., Loukissas, A. and Vahdat, A., 2008. A scalable, commodity data center
network architecture. ACM SIGCOMM Computer Communication Review, 38(4), pp.63-74.
3. Greenberg, A., Hamilton, J., Maltz, D.A. and Patel, P., 2008. The cost of a cloud: research
problems in data center networks. ACM SIGCOMM computer communication review, 39(1),
pp.68-73.
4. Guo, C., Lu, G., Li, D., Wu, H., Zhang, X., Shi, Y., Tian, C., Zhang, Y. and Lu, S., 2009.
BCube: a high performance, server-centric network architecture for modular data centers.
ACM SIGCOMM Computer Communication Review, 39(4), pp.63-74.
5. G. Kecskemeti (2015): DISSECT-CF: A simulator to foster energy-aware scheduling in
infrastructure clouds. Simulation Modelling Practice and Theory, 58 :188-218.
6. Rodrigo N. Calheiros et al. CloudSim: A Toolkit for Modeling and Simulation of Cloud
Computing Environments and Evaluation of Resource Provisioning Algorithms. Software:
Practice and Experience (SPE), Volume 41, Number 1, Pages 23-50. January 2011.
Extenuating Circumstances
If something serious happens that means that you will not be able to complete this assignment, you
need to contact the module leader as soon as possible. There are a number of things that can be
done to help, such as extensions, waivers and alternative assessments, but we can only arrange
this if you tell us. To ensure that the system is not abused, you will need to provide some evidence
of the problem.
More guidance is available at https://www.ljmu.ac.uk/about-us/public-information/student-
regulations/guidance-policy-and-process
Any coursework submitted late without the prior agreement of the module leader will
receive 0 marks.
Academic Misconduct
The University defines Academic Misconduct as ‘any case of deliberate, premeditated cheating,
collusion, plagiarism or falsification of information, in an attempt to deceive and gain an unfair
advantage in assessment’. This includes attempting to gain marks as part of a team without
making a contribution. The Faculty takes Academic Misconduct very seriously and any suspected
cases will be investigated through the University’s standard policy (https://www.ljmu.ac.uk/about-
us/public-information/student-regulations/appeals-and-complaints). If you are found guilty, you
may be expelled from the University with no award.
It is your responsibility to ensure that you understand what constitutes Academic
Misconduct and to ensure that you do not break the rules. If you are unclear about what is
required, please ask.
For more information you are directed to following the University web pages:
• Information regarding academic misconduct: https://www.ljmu.ac.uk/about-us/public-
information/student-regulations/appeals-and-complaints
• Information on study skills: https://www2.ljmu.ac.uk/studysupport/
• Information regarding referencing: