Auto recovery;
error handling mechanism;
degradation;
perfect/imperfect repair/debugging;
mean sojourn time;
mean time to system failure;
cost;
regenerative stochastic process;
D O I:
10.1080/09720510.2008.10701324
中图分类号:
O21 [概率论与数理统计];
C8 [统计学];
学科分类号:
020208 ;
070103 ;
0714 ;
摘要:
Study of combined effects of imperfect repair/maintenance/debugging and failure on the dependability of a software system is in the core of the present paper. Faults/bugs may cause the degradation or failures to the software. Failures, categorized as minor, marginal and critical need the maintenance personnel for software repair. Measures of the dependability such as mean time to system failure (MTSF), availability, mean sojourn time, and busy periods of the maintenance personnel due to different failures are derived using the regenerative stochastic process. The expected profit/cost aspect is also taken into consideration. Numerical examples are given for illustrating the graphical behavior of Dependability measures, MTSF and availability.