An Fast Transparent Failover Scheme for Service Availability

Abstract

Fault tolerance schemes can be used to increase the availability and reliability of network service systems. One aspect of such schemes is failover - the reconfiguration of available resources and restoration of state required to continue providing the service despite the loss of some of the resources and corruption of parts of the state. Most of the techniques used for increasing the availability of network services do not provide fault tolerance for requests being processed at the time of server failure. Other schemes require deterministic servers or changes to the client. These limitations are unacceptable for many current and future network services. This paper proposes an efficient transparent failover scheme for service availability (SAFailover) to provide fault-tolerant network service that does not have the limitations mentioned above. The scheme is based on a hot standby backup server that maintains logs of requests and replies. The implementation includes modifications to the Linux kernel and to the Apache Web server, using their respective module mechanisms. The paper describe the implementation and present an evaluation of the impact of the backup scheme in terms of throughput, latency, and CPU processing cycles overhead. Experiment results indicate this scheme have short failover times and low overhead during fault-free operation