| Summary: | 碩士 === 國立成功大學 === 製造工程研究所碩博士班 === 94 === In recent years, the issue of “Software Aging” has been proposed and studied. The exhaustion of key resources causes the performance decline and failure of an application server and even worse the entire information application system.
Aiming at this issue, this research extends the original framework of Performance Evaluator (PEV), adopts the Platform-Based Design methodology, employs ARM Platform and Embedded Linux as the central techniques to develop an ARM-based performance evaluator. The developed PEV platform incorporates both the Simple Network Management Protocol (SNMP) and a remote network monitoring system, called Cacti.
Detection Module and Prediction Module are two kernel modules of the PEV. We propose a performance index to determine the performance status of application servers. The response time and fast Fourier transform are used to verify the accuracy of the Detection Module. In addition, the Prediction Module is constructed by using Lagrange polynomial. When Detection Module detects performance anomalies of application server, it will inform Prediction Module to predict the “Time to Failure” of the application server according to the residual available resources. Lastly, PEV provides the “Time to Failure” and relative messages to the application cluster service (APCS) to perform the failover of failed services before the upcoming system breakdown to achieve near-zero downtime services.
|
|---|
