We calculate nALT to be 25 using formula 3 to 7 , which is the minimum number of samples needed for our ALT test plan. Since this number is smaller than we obtained in our sampling, we satisfy the ALT assumptions for sample size. The fitting results for these four models are shown in Table IV. The probability density function pdf of Lognormal is shown in Equation 8. Figure 5. The criterion used The life characteristic for the Lognormal distribution to build the best-fit ranking is the log-likelihood function is its median value that is given by 9 Lk [28].
Table V gives the parameter estimation of lognormal model by the ML estimation method [23]. We use Equation 12 to estimate the underlying Figure 6 a presents the fit of the IPL-lognormal model estimated from the four stress levels, and for the normal condition. Figure 6 b shows the failure and stress relationships. The 5. As a result, we can obtain the MTTF 1.
Probability plot for the normal condition level 2. These results can be used to further schedule software rejuvenation, and thus improve the software availability, and reduce the maintenance costs. Normal Standardized residuals. We optimize the software rejuvenation trigger interval in order to maximize the system availability or Figure 8 shows this model. State 0 is the only available state. The distribution function for the duration w 0. This model is a semi-Markov process [28].
Also, the parameter E confidence interval, Figure 8. Rejuvenation model. Steady-state availability vs time to rejuvenation K K trigger, t0, assuming the Weibull time to failure distribution where, is shown in Figure 9. The optimal time to trigger f t t 0 , t t 0 , E! The average cost vs. This is a three-parameter model. The estimated time to rejuvenation t0 is shown in Figure Our results show only minor difference between the goodness of fit between the Weibull and the Lognormal.
We also compute the availability confidence interval and operational cost at the optimal rejuvenation trigger interval using the estimated parameters from our experiments. Grottke, L. Li, K. Vaidyanathan, K. Grottke, A. Nikora, K. Garg, A. Huang, C. Kinatla, N. Grottke, R. Matias, and K. Average cost vs. Jia, L. Zhao and K. In this paper, we develop experiments that simulate [7] E. We , pp. Based on the [9] K. Cassidy, K. Gross, and A. International conference on memory consumption rate was selected as the acceleration dependable systems and networks, pp.
Secondly, the IPL-lognormal model was built to [10] X. Zhang and H. Bao, X. Sun, K. A considerable reduction in experimentation [13] T. Dohi, K. Goseva-Popstojanova, and K. Four stress non-parametric algorithms to estimate the optimal software levels and 7 replications are used at each stress level. Puliafito, M. Telek,, K. Silva, H. Madeira, J. Alonso, J. Berral, R. Alonso, L. Silva, A. Andrzejak, P. Silva, J. Torres, [23] B. Wrong email address.
You're going to remove this assignment. Are you sure? Yes No. Keywords Aging Stress Memory management Acceleration Java Web servers software aging accelerated life tests memory leaks optimal software rejuvenation semi-Markov process Aging Stress Memory management Acceleration Java Web servers software aging accelerated life tests memory leaks optimal software rejuvenation semi-Markov process. Additional information Data set: ieee.
Publisher IEEE. You have to log in to notify your friend by e-mail Login or register account. Download to disc. High contrast On Off. Close window. Assign to yourself.
We study the application failures caused by memory leaks, using the accelerated life tests method. In our experiments, the memory consumption rate is selected as the acceleration factor, and an IPL-lognormal model is used to estimate the time to failure at each acceleration level.
Subsequently, the estimate of the time to failure distribution at normal condition is obtained. Our acceleration experimental results based on the IPL-lognormal model show that it can be used to greatly reduce the cost to obtain the time to failure at normal level, which can be used in scheduling software rejuvenation.
0コメント