Search results

1 – 1 of 1
Article
Publication date: 15 April 2020

Chandra Shekhar, Amit Gupta, Madhu Jain and Neeraj Kumar

The purpose of this paper is to present a sensitivity analysis of fault-tolerant redundant repairable computing systems with imperfect coverage, reboot and recovery process.

Abstract

Purpose

The purpose of this paper is to present a sensitivity analysis of fault-tolerant redundant repairable computing systems with imperfect coverage, reboot and recovery process.

Design/methodology/approach

In this investigation, the authors consider the computing system having a finite number of identical working units functioning simultaneously with the provision of standby units. Working and standby units are prone to random failure in nature and are administered by unreliable software, which is also likely to unpredictable failure. The redundant repairable computing system is modeled as a Markovian machine interference problem with exponentially distributed failure rates and service rates. To excerpt the failed unit from the computing system, the system either opts randomized reboot process or leads to recovery delay.

Findings

Transient-state probabilities have been determined with which the authors develop various reliability measures, namely reliability/availability, mean time to failure, failure frequency, and so on, and queueing characteristics, namely expected number of failed units, the throughput of the system and so on, for the predictive purpose. To spectacle the practicability of the developed model, a numerical simulation, sensitivity analysis and so on for different parameters have also been done, and the results are summarized in the tables and graphs. The transient results are helpful to analyze the developing model of the system before having the stability of the system. The derived measures give direct insights into parametric decision-making.

Social implications

The conclusion has been drawn, and future scope is remarked. The present research study would help system analyst and system designer to make a better choice/decision in order to have the economical design and strategy based on the desired mean time to failure, reliability/availability of the systems and other queueing characteristics.

Originality/value

Different from previous investigations, this studied model provides a more accurate assessment of the computing system compared to uncertain environments based on sensitivity analysis.

Details

International Journal of Quality & Reliability Management, vol. 37 no. 6/7
Type: Research Article
ISSN: 0265-671X

Keywords

1 – 1 of 1