
Availability Modelling of Cluster-Based System with Software Aging and Optional Rejuvenation Policy
Author(s) -
Richa Sharma,
Gireesh Kumar
Publication year - 2019
Publication title -
cybernetics and information technologies
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.272
H-Index - 17
eISSN - 1314-4081
pISSN - 1311-9702
DOI - 10.2478/cait-2019-0038
Subject(s) - downtime , rejuvenation , computer science , reliability engineering , software , reboot , cluster analysis , software maintenance , software system , high availability , cluster (spacecraft) , operating system , engineering , gerontology , medicine , machine learning
This investigation deals with modeling and availability analysis of cluster-based system inflicted with software aging. Software aging is a phenomenon in which a software system shows performance degradation with time and finally results in software failures. To cope up with this phenomenon, rejuvenation is an innovative concept to recover from software failures. As failures occur, server has the option either to take essential rejuvenation with probability p or may opt for optional rejuvenation with complementary probability q . To achieve high availability of the system, the concept of clustering is also taken into consideration. In this study, restart, reboot and standby concept is used for reducing the downtime cost. The sensitivity analysis of different parameters on system availability has been examined numerically. By integrating clustering, software aging and rejuvenation, the researchers intended to increase the availability and decrease the down time.