Percorrer por autor "Pinheiro, Henrique Sousa"
A mostrar 1 - 1 de 1
Resultados por página
Opções de ordenação
- An eventually perfect failure detector in a high-availability scenarioPublication . Pinheiro, Henrique Sousa; Lopes, Rui Pedro; Campiolo, RodrigoModern-day distributed systems have been increasing in complexity and dynamism due to the heterogeneity of the system execution environment, different network technologies, online repairs, frequent updates and upgrades, and the addition or removal of system components. Such complexity has elevated the operational and maintenance costs and triggered efforts to reduce it while improving its reliability. Availability is the ratio of uptime to total time of a system. A High Available system, or systems with at least 99.999% of Availability, imposes a challenge to maintain such levels of uptime. Prior work shows that by using system state monitoring and fault management with failure detectors it is possible to increase system availability. The main objective of this work is to develop an Eventually Perfect Failure Detector to improve a database system Availability through fault-tolerance methods. Such a system was developed and tested in a proposed High-Availability database access infrastructure. Final results have shown that is possible to achieve performance and availability improvements by using, respectively, replication and a failure detector.