隨著科技的進步，資訊應用系統對企業或是個人用戶來說佔有舉足輕重的地位，一旦資訊應用系統可靠度不佳，對使用者來說會遭受到嚴重的損失。近年來 “Software aging” 的現象被提出：資訊應用系統因關鍵資源耗盡造成效能衰減及失效。針對此現象，本研究設計及實作出具預測應用伺服器失效能力之效能評估器 (Performance Evaluator, PEV)，我們利用模糊理論 (Fuzzy Theory) 及時間序列方法 (Time Series Analysis) 建構出此評估器中的兩個核心架構： “效能偵測模組”及 “失效預測模組”；其分別具備了監控應用伺服器效能異常偵測及失效預測之能力。效能偵測模組具有監控及偵測應用伺服器健康之能力，如果其偵測到應用伺服器不健康，則會通知失效預測模組預測其應用伺服器之失效時間 (Time to Failure)。另外，我們使用馬可夫鏈 (Markov Chain) 驗證此效能偵測架構之可行性。最後本效能評估器可提供失效時間等訊息給APCS (Application Cluster Service, APCS)模組，使其可提前將欲失效的服務做轉移，俾達成服務不中斷的境界。

With the progress of the technology, information application systems become more and more important to enterprises, and even individuals’ daily life. Once an application server is not reliable, it may cause great loss to the users. In recent years, the issue of “software aging” has been proposed and studied. The exhaustion of key resources causes the performance decline and failure of an application server and even worse the entire information application system. Aiming at this issue, this work designs and implements a Performance Evaluator (PEV), which possesses the capability of failure prognosis. We apply the fuzzy theory and time series analysis to construct two main kernel schemes of the proposed PEV, the performance detection and failure prediction modules. The detection module has the capability of monitoring and detecting the healthy condition of each application server. If a sickness condition of an application server is detected, the detection module will request the prediction module to predict the time to failure of this unhealthy application server. In addition, the feasibility of the detection scheme is verified with the theory of Markov Chain. This work further demonstrates that the information of time to failure can be used by the application cluster service scheme (APCS) to perform failure process before the breakdown of this unhealthy application server such that near-zero downtime services are assured.

參考文獻
[1] G. J. Gray, “Why Do Computers Stop and What Can Be Done about It,” Proc. of 5th Symp on Reliability in Distributed Software and Database Systems, pp. 3-12, January. 1986.
[2] A. Avritzer and E. J. Weyuker, “Monitoring Smoothly Degrad-ing Systems for Increased Dependability,” Empirical Software Eng. Journal, Vol 2, no. 1, pp 59–77, 1997.
[3] L. Bernstein, text of seminar delivered at the University Learning Center, George Mason University, January. 1996.
[4] A. Bobbio, A. Sereno and C. Anglano, “Fine Grained Soft-ware Degradation Models for Optimal Rejuvenation Policies,” Performance Evaluation, 46, pp. 45–62, 2001.
[5] V. Castelli, R.E. Harper, P. Heidelberger, S.W. Hunter, K.S.Trivedi, K. Vaidyanathan and W.P. Zeggert, “Proactive Management of Software Aging,” IBM JRD, Vol 45, no. 2, pp.311–332, Mar. 2001.
[6] K. Cassidy, K. Gross and A. Malekpour, “Advanced Pattern Recognition for Detection of Complex Software Aging in On-line Transaction Processing Servers,” Proc. DSN 2002, Washington D.C., June. 2002.
[7] T. Dohi, K. Goseva-Popstojanova and K.S. Trivedi, “Statisti-cal Non-parametric Algorithms to Estimate the Optimal Soft-ware Rejuvenation Schedule,” Proc. of PRDC 2000, LosAn-geles, CA, Dec. 2000.
[8] S. Garg, A. Puliafito and K.S. Trivedi, “Analysis of Soft-ware Rejuvenation using Markov Regenerative Stochastic Petri Net,” Proc. of ISSRE 1995, Toulouse, France, Oct. 1995.
[9] S. Garg, A. van Moorsel, K. Vaidyanathan and K.S. Trivedi,“A Methodology for Detection and Estimation of Software Aging,” Proc. of ISSRE 1998, Paderborn, Germany, Nov. 1998.
[10] S. Garg, A. Puliafito, M. Telek and K.S. Trivedi, “Analysis of Preventive Maintenance in Transaction Processing Systems,” IEEE Trans. on Computers, 47(1), pp. 96–107, Jan. 1998.
[11] J. Gray and D. P. Siewiorek, “High-availability Computer Systems,” IEEE Computer, 24(9), pp. 39–48, Sep. 1991.
[12] Y. Huang, C. Kintala, N. Kolettis and N.D. Fulton, “Software Rejuvenation: Analysis, Module and Applications,” Proc. of FTCS-25, Pasadena, CA, Jun. 1995.
[13] IBM Netfinity Director Software Rejuvenation - White Paper, IBM Corp., Research Triangle Park, NC, Jan. 2001.
[14] Y. Liu, Y. Ma, J. J. Han, H. Levendel and K.S. Trivedi, “Modeling and Analysis of Software Rejuvenation in Cable Modem Termination System,” in Proc. of ISSRE 2002, Annapolis, MD, Nov. 2002.
[15] E. Marshall, “Fatal Error: How Patriot Overlooked a Scud,” Science, pp. 1347, Mar. 13 1992.
[16] S. Pfening, S. Garg, A. Puliafito, M. Telek and K.S. Trivedi, “Optimal Rejuvenation for Tolerating Soft Failures”, Performance Evaluation, pp. 491-506, 27/28, 1996.
[17] P.K. Sen, “Estimates of the Regression Coefficient Based on Kendall’s Tau,” Journal of the American Statistical Association, 63:1379–1389, 1968.
[18] A.T. Tai, S.N. Chau, L. Alkalaj and H. Hecht, “On-Board Preventive Maintenance: Analysis of Effectiveness and Optimal Duty Period”, Proc. 3rd Intl. Workshop on Object Ori-ented Real-time Dependable Systems, Newport Beach, CA, Feb. 1997.
[19] K. Vaidyanathan and K.S. Trivedi, “A Measurement-based Model for Estimation of Resource Exhaustion in Operational Software Systems,” Proc. Of ISSRE 1999, Boca Raton, FL, Nov. 1999.
[20] K. Vaidyanathan, R.E. Harper, S.W. Hunter and K.S. Trivedi,“Analysis and Implementation of Software Rejuvenation in Cluster Systems,” Proc. of ACM SIGMETRICS 2001/Performance 2001, Cambridge, MA, Jun. 2001.
[21] K. Vaidyanathan, D. Selvamuthu and K.S. Trivedi, “Analysis of Inspection-Based Preventive Maintenance in Operational Software Systems,” Proc. of SRDS 2002, Oct. 2002.
[22] HTTP Server, Apache Corporation. [Online]. Available: http://httpd.apache.org/docs/stopping.html
[23] IMS Vision, Center for Intelligent Maintenance Systems. [Online]. Available: http://wumrc.engin.umich.edu/ims/?page=home
[24] R. Gamache, R. Short, and M. Massa, “Windows NT Clustering Service,” Computer, vol. 31, no. 10, pp.55-62, Oct. 1998.
[25] W. Vogels, D. Dumitriu, K. Birman, R. Gamache, M. Massa, R. Short, J. Vert, J. Barrera, and J. Gray, “The Design and Architecture of the Microsoft Cluster Service—A Practical Approach to High-Availability and Scalability,”in Proc. 28th Symp. Fault-Tolerant Computing, CS Press, 1998, pp. 422-431.
[26] Matrix HA/Server, PolyServe Corporation. [Online]. Available: http://www.polyserve.com/
[27] K. Shen, T. Yang, and L. Chu, “Clustering Support and Replication Management for Scalable Network Services,” IEEE Transactions on Parallel and Distributed Systems, vol. 14, no. 11, pp.1168-1179, Nov. 2003.
[28] F.-T. Cheng, H.-C. Yang, and C.-Y. Tsai, “Developing a Service Management Scheme for Semiconductor Factory Management Systems,” IEEE Robotics and Automation Management, vol. 11, no. 1, pp. 26-40, March 2004.
[29] M. Sullivan and R. Chillarege, “Software Defects and their Impact on System Availability — A Study of Field Failures in Operating Systems,” Proc. of FTCS-21, Montreal, Canada, Jun 1991.
[30] UML, OMG's 1997 Press Releases. [Online]. Available: http://www.omg.org/news/pr97.htm
[31] Perfmon, Microsoft Corporation. [Online]. Available: http://msdn.microsoft.com/
[32] Performance Monitoring, Microsoft Corporation. [Online]. Available: http://www.microsoft.com/windows2000/techinfo/reskit/en-us/default.asp?url=/windows2000/techinfo/reskit/en-us/core/fneb_mon_bhod.asp/
[33] MS Web Application Stress Tool, Microsoft Corporation. [Online]. Available: http://www.microsoft.com/technet/archive/itsolutions/intranet/downloads/webstres.mspx/
[34] 王淵海，統計學，文笙書局，台灣，台北，民國八十八年五月.
[35] Kishor S. Trivedi, Probability and Statistics with Reliability, Queuing and Computer Science Applications (2nd ed.), ISBN0-471-33341-7, New York：John Wiley & Sons, 2002.
[36] Borland J++ Builder 6.0, Borland Corporation. [Online]. Available: http://www.borland.com/cbuilderx/
[37] Fuzzytech, INFORM Gmbh Corporation. [Online]. Available: http://www.inform-ac.com/fuzzy/en/fuzzy.htm
[38] 吳尚倫 「具錯誤後轉移能力之應用程式叢集服務」，國立成功大學製造工程研究所碩士論文2004.
[39] SPSS, SPSS Corporation. [Online]. Available: http://www.spss.com/