A Dynamic Anomaly Detection Approach Based on Permutation Entropy for Predicting Aging-Related Failures

Shuguang Wang; Minyan Lu; Shiyi Kong; Jun Ai

doi:10.3390/e22111225

A Dynamic Anomaly Detection Approach Based on Permutation Entropy for Predicting Aging-Related Failures

Entropy ◽

10.3390/e22111225 ◽

2020 ◽

Vol 22 (11) ◽

pp. 1225

Author(s):

Shuguang Wang ◽

Minyan Lu ◽

Shiyi Kong ◽

Jun Ai

Keyword(s):

Performance Metrics ◽

Distributed Database ◽

Dynamic State ◽

Permutation Entropy ◽

Software Aging ◽

Multi Scale ◽

Detection Approach ◽

State Changes ◽

Distributed Database System ◽

Aging Indicator

Software aging is a phenomenon referring to the performance degradation of a long-running software system. This phenomenon is an accumulative process during execution, which will gradually lead the system from a normal state to a failure-prone state. It is a crucial challenge for system reliability to predict the Aging-Related Failures (ARFs) accurately. In this paper, permutation entropy (PE) is modified to Multidimensional Multi-scale Permutation Entropy (MMPE) as a novel aging indicator to detect performance anomalies, since MMPE is sensitive to dynamic state changes. An experiment is set on the distributed database system Voldemort, and MMPE is calculated based on the collected performance metrics during execution. Finally, based on MMPE, a failure prediction model using the machine learning method to reveal the anomalies is presented, which can predict failures with high accuracy.

Download Full-text

Performance Evaluation of Distributed Database Systems

INTERNATIONAL JOURNAL OF COMPUTERS & TECHNOLOGY ◽

10.24297/ijct.v13i9.2383 ◽

2014 ◽

Vol 13 (9) ◽

pp. 4859-4867

Author(s):

Khaled Saleh Maabreh

Keyword(s):

Performance Evaluation ◽

System Performance ◽

Operation Mode ◽

Database Systems ◽

Distributed Database ◽

Evaluation Study ◽

Distributed Database Systems ◽

Different Types ◽

Distributed Database System ◽

Distributed Database Management

Distributed database management systems manage a huge amount of data as well as large and increasingly growing number of users through different types of queries. Therefore, efficient methods for accessing these data volumes will be required to provide a high and an acceptable level of system performance. Â Data in these systems are varying in terms of types from texts to images, audios and videos that must be available through an optimized level of replication. Distributed database systems have many parameters like data distribution degree, operation mode and the number of sites and replication. These parameters have played a major role in any performance evaluation study. This paper investigates the main parameters that may affect the system performance, which may help with configuring the distributed database system for enhancing the overall system performance.

Download Full-text