Sierpinski triangle based data-center network architecture in cloud computing / Qi Han
This thesis reports on the research to develop of a data center network (DCN) architecture to solve the problem of network performance in cloud-oriented data centers. Computational clouds are increasingly becoming popular for the provisioning of computing resources and service on demand basis. A...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/6154/ http://studentsrepo.um.edu.my/6154/4/qi_han.pdf |
| _version_ | 1848773086727372800 |
|---|---|
| author | Qi, Han |
| author_facet | Qi, Han |
| author_sort | Qi, Han |
| building | UM Research Repository |
| collection | Online Access |
| description | This thesis reports on the research to develop of a data center network (DCN) architecture
to solve the problem of network performance in cloud-oriented data centers.
Computational clouds are increasingly becoming popular for the provisioning of computing
resources and service on demand basis. A DCN is an important component of data
centers that consists of a large number of hosted servers and switches connected with high
speed communication links. As a backbone in data centers, a DCN enables the deployment
of resources centralization and on-demand access of the information and services
of data centers to users. In recent years, the scale of the DCN has constantly increased
with the widespread use of cloud-oriented services and applications configured over virtual
machines (VMs), and the unprecedented amount of data delivery in/between data
centers, whereas the traditional DCN tree-based architecture lacks aggregate bandwidth,
scalability and cost effectiveness for coping with the increasing demands of tenants in
accessing the services of cloud-oriented data centers. To solve this problem, the method
developed in this research is used to mitigate the aggregation throughput and improve
the network performance of DCN by using a novel DCN architecture. The proposed
method, called Sierpinski Triangle Based (STB) DCN architecture, is developed on the
basis of the well-know Sierpinski triangle fractal to mitigate throughput bottleneck in
aggregate layers as accumulated in tree-based structure. STB is a fault-tolerant architecture
which provides at least two parallel paths for each pair of servers. It also supports
various bandwidth-intensive applications by providing high network throughput for allto-
all traffic. STB architecture was implemented in a real cloud data center environment
and evaluated in Network Simulator 2 (NS2) simulation. The performance of STB architecture
is validated by comparing the results with the traditional tree-based, and DCell
DCN architectures. Theoretical analysis and implementation experiences verify that the
iii
proportion of server to entire nodes in STB is same with DCell but higher than that of
tree-based architecture. The average shortest path length is restricted between 5.0 and
6.7, when node failure proportion remains between 0.02 and 0.2, shorter than DCell in
a 4-level architecture. The results of the experiment also show that the STB architecture
has higher throughout than both traditional tree-based and DCell architectures from the
scale of 12 to 363 servers with/without link failure happens. From the results of both simulation
and experiment in actual devices, we speculate that STB still can achieve better
network performance in throughput, server utilization, average shortest path length than
DCell and tree-based architectures in real large-scale cloud-oriented DCN. |
| first_indexed | 2025-11-14T13:36:49Z |
| format | Thesis |
| id | um-6154 |
| institution | University Malaya |
| institution_category | Local University |
| last_indexed | 2025-11-14T13:36:49Z |
| publishDate | 2016 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | um-61542019-09-10T23:01:14Z Sierpinski triangle based data-center network architecture in cloud computing / Qi Han Qi, Han QA75 Electronic computers. Computer science QA76 Computer software This thesis reports on the research to develop of a data center network (DCN) architecture to solve the problem of network performance in cloud-oriented data centers. Computational clouds are increasingly becoming popular for the provisioning of computing resources and service on demand basis. A DCN is an important component of data centers that consists of a large number of hosted servers and switches connected with high speed communication links. As a backbone in data centers, a DCN enables the deployment of resources centralization and on-demand access of the information and services of data centers to users. In recent years, the scale of the DCN has constantly increased with the widespread use of cloud-oriented services and applications configured over virtual machines (VMs), and the unprecedented amount of data delivery in/between data centers, whereas the traditional DCN tree-based architecture lacks aggregate bandwidth, scalability and cost effectiveness for coping with the increasing demands of tenants in accessing the services of cloud-oriented data centers. To solve this problem, the method developed in this research is used to mitigate the aggregation throughput and improve the network performance of DCN by using a novel DCN architecture. The proposed method, called Sierpinski Triangle Based (STB) DCN architecture, is developed on the basis of the well-know Sierpinski triangle fractal to mitigate throughput bottleneck in aggregate layers as accumulated in tree-based structure. STB is a fault-tolerant architecture which provides at least two parallel paths for each pair of servers. It also supports various bandwidth-intensive applications by providing high network throughput for allto- all traffic. STB architecture was implemented in a real cloud data center environment and evaluated in Network Simulator 2 (NS2) simulation. The performance of STB architecture is validated by comparing the results with the traditional tree-based, and DCell DCN architectures. Theoretical analysis and implementation experiences verify that the iii proportion of server to entire nodes in STB is same with DCell but higher than that of tree-based architecture. The average shortest path length is restricted between 5.0 and 6.7, when node failure proportion remains between 0.02 and 0.2, shorter than DCell in a 4-level architecture. The results of the experiment also show that the STB architecture has higher throughout than both traditional tree-based and DCell architectures from the scale of 12 to 363 servers with/without link failure happens. From the results of both simulation and experiment in actual devices, we speculate that STB still can achieve better network performance in throughput, server utilization, average shortest path length than DCell and tree-based architectures in real large-scale cloud-oriented DCN. 2016 Thesis NonPeerReviewed application/pdf http://studentsrepo.um.edu.my/6154/4/qi_han.pdf Qi, Han (2016) Sierpinski triangle based data-center network architecture in cloud computing / Qi Han. PhD thesis, University of Malaya. http://studentsrepo.um.edu.my/6154/ |
| spellingShingle | QA75 Electronic computers. Computer science QA76 Computer software Qi, Han Sierpinski triangle based data-center network architecture in cloud computing / Qi Han |
| title | Sierpinski triangle based data-center network architecture in cloud computing / Qi Han |
| title_full | Sierpinski triangle based data-center network architecture in cloud computing / Qi Han |
| title_fullStr | Sierpinski triangle based data-center network architecture in cloud computing / Qi Han |
| title_full_unstemmed | Sierpinski triangle based data-center network architecture in cloud computing / Qi Han |
| title_short | Sierpinski triangle based data-center network architecture in cloud computing / Qi Han |
| title_sort | sierpinski triangle based data-center network architecture in cloud computing / qi han |
| topic | QA75 Electronic computers. Computer science QA76 Computer software |
| url | http://studentsrepo.um.edu.my/6154/ http://studentsrepo.um.edu.my/6154/4/qi_han.pdf |