In order to achieve high availability the idea of cluster systems has been commonly used. In this paper a two node cluster system is considered in which software and hardware failures occur. Hardware failures are permanent and hence they are counteracted by a repair or replacement. On the other side, software failures require a reboot. A reboot though is a reactive action that can be taken when software failures due to degradation are observed. Software rejuvenation, a proactive kind of maintenance, is used to counteract software degradation called "software aging". In the cluster system presented, software rejuvenation is performed when the node's software experiences degradation in order to avoid an unscheduled reboot. The system is modeled under continuous time Markov chains (CTMC). Two different clustering techniques are studied and compared in order to determine the best configuration which provides the highest level of availability
Description
IEEE Xplore Abstract - Applying Software Rejuvenation in a Two Node Cluster System for High Availabilit
%0 Conference Paper
%1 koutras2006applying
%A Koutras, V.P.
%A Platis, A.N.
%B Dependability of Computer Systems, 2006. DepCos-RELCOMEX '06. International Conference on
%D 2006
%K rejuvenation
%P 175-182
%R 10.1109/DEPCOS-RELCOMEX.2006.7
%T Applying Software Rejuvenation in a Two Node Cluster System for High Availabilit
%U http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=4024047
%X In order to achieve high availability the idea of cluster systems has been commonly used. In this paper a two node cluster system is considered in which software and hardware failures occur. Hardware failures are permanent and hence they are counteracted by a repair or replacement. On the other side, software failures require a reboot. A reboot though is a reactive action that can be taken when software failures due to degradation are observed. Software rejuvenation, a proactive kind of maintenance, is used to counteract software degradation called "software aging". In the cluster system presented, software rejuvenation is performed when the node's software experiences degradation in order to avoid an unscheduled reboot. The system is modeled under continuous time Markov chains (CTMC). Two different clustering techniques are studied and compared in order to determine the best configuration which provides the highest level of availability
@inproceedings{koutras2006applying,
abstract = {In order to achieve high availability the idea of cluster systems has been commonly used. In this paper a two node cluster system is considered in which software and hardware failures occur. Hardware failures are permanent and hence they are counteracted by a repair or replacement. On the other side, software failures require a reboot. A reboot though is a reactive action that can be taken when software failures due to degradation are observed. Software rejuvenation, a proactive kind of maintenance, is used to counteract software degradation called "software aging". In the cluster system presented, software rejuvenation is performed when the node's software experiences degradation in order to avoid an unscheduled reboot. The system is modeled under continuous time Markov chains (CTMC). Two different clustering techniques are studied and compared in order to determine the best configuration which provides the highest level of availability},
added-at = {2014-10-15T19:44:52.000+0200},
author = {Koutras, V.P. and Platis, A.N.},
biburl = {https://www.bibsonomy.org/bibtex/236572352bc51ffe92e4112eb14efb219/avail_map_stud},
booktitle = {Dependability of Computer Systems, 2006. DepCos-RELCOMEX '06. International Conference on},
description = {IEEE Xplore Abstract - Applying Software Rejuvenation in a Two Node Cluster System for High Availabilit},
doi = {10.1109/DEPCOS-RELCOMEX.2006.7},
interhash = {b8acf59d0a4f6a72b93044a2b7a2acd6},
intrahash = {36572352bc51ffe92e4112eb14efb219},
keywords = {rejuvenation},
month = may,
pages = {175-182},
timestamp = {2014-10-15T19:44:52.000+0200},
title = {Applying Software Rejuvenation in a Two Node Cluster System for High Availabilit},
url = {http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=4024047},
year = 2006
}