在網路快速發展的現今，網路的需求與日俱增。然而，傳統網路架構難以應付陡然遽增的網路流量，軟體定義網路(Software-Defined Network, SDN)因此順勢而生。在軟體定義網路中，網路的管理權限將由控制層的 SDN 控制器(SDN Controller)負責，採用了集中管理的方式。關於集中管理的方式，在某些只有單一 SDN 控制器的情境中，若是 SDN 控制器故障，將會造成整個網路的癱瘓，進而造成極大的損失。於是，便有了多個 SDN 控制器共同管理網路的方式，但在多個 SDN 控制器共同管理網路的情況下，需要各個 SDN 控制器彼此達成一致性的結果，所以，需要共識機制來達成。雖然有多個 SDN 控制器共同進行管理，但是在錯誤的控制器數量超過總控制器數量一半的情況下，仍然會導致整個網路失去控制，因此，仍然需要一個妥善的 SDN 控制器恢復機制來保證網路的可用性。但在現有的共識與恢復機制中，不是恢復時間過久導致整個系統的反應時間變長，就是需要額外佈署機台遠端監控，造成額外的通訊成本以及安全疑慮。因此，本碩士論文提出了一個自我監督的機制，可以縮短整個系統的反應時間，且不須額外的通訊成本，同時避免了可能的資安問題。透過模擬，我們呈現所提出的自我監督機制的確能夠有效地改善整個系統的效能。

Nowadays, the demand for internet is growing rapidly as the network technology evolves. However, the traditional network architectures are unable to cope with the sudden increase of network traffic, the software-defined network (SDN) is then come up with. In SDN, the SDN controller at the control level will be responsible for network management with a centralized manner.
Of course, a risk for this centralized manner exists inevitably. In the scenario when only a single SDN controller is assumed but it fails, the entire network will be broken down, causing losses. A solution with multiple SDN controllers to jointly manage the network is then devised. For reaching a consistent result with each SDN controller, a consensus mechanism is required for sure. When the number of failed SDN controllers exceeds a half of the total SDN controllers, it still makes the entire network malfunction. Therefore, a proper SDN controller recovery mechanism is a must to ensure the network availability. For the existing consensus and recovery mechanisms, either a longer recovery time incurs or a remote monitoring agent is required, causing an additional communication cost or even security concerns. Therefore, a self-monitoring mechanism which can reduce the response time of the entire system without an additional communication cost and possible security concerns is proposed in this thesis. Through simulations, we show that the proposed self-monitoring mechanism can effectively improve the performance of the system.

[1] I. F. Akyildiz, S.-C. Lin, and P. Wang, “Wireless software-defined networks
(W-SDNs) and network function virtualization (NFV) for 5G cellular systems:
An overview and qualitative evaluation,” Computer Networks, vol. 93, pp. 66
– 79, Dec. 2015.
[2] J. Medved, R. Varga, A. Tkacik, and K. Gray, “OpenDaylight: Towards a
model-driven SDN controller architecture,” in Proc. of IEEE International
Symposium on a World of Wireless, Mobile and Multimedia Networks 2014,
pp. 1–6, Jun. 2014.
[3] P. Berde, M. Gerola, J. Hart, Y. Higuchi, M. Kobayashi, T. Koide, B. Lantz,
B. O’Connor, P. Radoslavov, W. Snow, et al., “ONOS: towards an open, distributed
SDN OS,” in Proc. of the third workshop on Hot topics in software
defined networking, pp. 1–6, Aug. 2014.
[4] S. Gilbert and N. Lynch, “Brewer’s conjecture and the feasibility of consistent,
available, partition-tolerant web services,” Acm Sigact News, vol. 33, no. 2,
pp. 51–59, 2002.
[5] Y. Liu, Y. Ma, J. J. Han, H. Levendel, and K. S. Trivedi, “A proactive approach
towards always-on availability in broadband cable networks,” Elsevier Computer
Communications, vol. 28, pp. 51–64, Jan. 2005.
[6] E. Sakic and W. Kellerer, “Response time and availability study of RAFT consensus
in distributed SDN control plane,” IEEE Transactions on Network and
Service Management, vol. 15, pp. 304–318, Aug. 2017.
[7] V. Saini, “Consensuspedia: An encyclopedia of 30+ consensus
algorithms,” June 2018. [Online]. Available: https:// hackernoon.
com/ consensuspedia-an-encyclopedia-of-29-consensus-algorithmse9c4b4b7d08f?
gi=b1de083eb98e.
[8] D. Mingxiao, M. Xiaofeng, Z. Zhe, W. Xiangwei, and C. Qijun, “A review on
consensus algorithm of blockchain,” in Proc. IEEE International Conference
on Systems, Man, and Cybernetics (SMC), pp. 2567–2572, Oct. 2017.
[9] S. J. Alsunaidi and F. A. Alhaidari, “A survey of consensus algorithms for
blockchain technology,” in Proc. International Conference on Computer and
Information Sciences (ICCIS), pp. 1–6, May 2019.
[10] M. Burrows, “The Chubby lock service for loosely-coupled distributed systems,”
in Proc. of the 7th symposium on Operating systems design and implementation,
pp. 335–350, Sep. 2006.
[11] L. LAMPORT, “The part-time parliament,” ACM Transactions on Computer
Systems, vol. 16, pp. 133–169, May 1998.
[12] I. Moraru, D. G. Andersen, and M. Kaminsky, “There is more consensus in
egalitarian parliaments,” in Proc. of the Twenty-Fourth ACM Symposium on
Operating Systems Principles, pp. 358–372, Nov. 2013.
[13] D. Ongaro and J. Ousterhout, “In search of an understandable consensus
algorithm,” in Annual Technical Conference, pp. 305–319, May 2014.
[14] D. Suh, S. Jang, S. Han, S. Pack, T. Kim, and J. Kwak, “On performance of
OpenDaylight clustering,” in Proc. IEEE NetSoft Conference and Workshops
(NetSoft), pp. 407–410, 2016.
[15] D. Suh, S. Jang, S. Han, S. Pack, M.-S. Kim, T. Kim, and C.-G. Lim, “Toward
highly available and scalable software defined networks for service
providers,” IEEE Communications Magazine, vol. 55, pp. 100–107, Apr.
2017.
[16] G. Bolch, S. Greiner, H. De Meer, and K. S. Trivedi, Queueing networks and
Markov chains: modeling and performance evaluation with computer science
applications. 2006.
[17] W. H. Sanders and J. F. Meyer, “Stochastic activity networks: Formal definitions
and concepts,” in School organized by the European Educational
Forum, pp. 315–343, Sep. 2000.
[18] H. Howard, M. Schwarzkopf, A. Madhavapeddy, and J. Crowcroft, “Raft
refloated: Do we have consensus?,” Operating Systems Review, vol. 49,
pp. 12–21, Jan. 2015.
[19] L. Kleinrock, Queueing systems. Volume I: theory. 1975.
[20] J. F. Shortle, J. M. Thompson, D. Gross, and C. M. Harris, Fundamentals of
queueing theory, vol. 399. 2018.
[21] R. De Smet, “Simulation modeling and analysis: Averill m. law and w. david
kelton mcgraw-hill,” 1993.