Data Center Handbook, 2nd Edition

29RELIABILITY ENGINEERING FOR DATA CENTER INFRASTRUCTURES

Malik Megdiche

Schneider Electric, Eybens, France

29.1 INTRODUCTION

Reliability engineering is defined as the science of failure. The first issues and reliability concepts appeared at the beginning of the twentieth century while engineering systems in railways, power systems, and aircraft applications. The main theories and methods have been developed for critical applications such as military, aerospace, and nuclear power plant. Nowadays, reliability engineering is widely used in many areas.

Reliability engineering uses equipment reliability statistics, probability theories, system functional analysis, and dysfunctional analysis to set requirements, measure or predict reliability, identify system weakness points, and propose improvements of the system.

Various reliability engineering techniques are used in reliability engineering:

• Equipment reliability analysis:
  • Field experience reliability statistics
  • Reliability testing
  • Accelerated life testing
• System reliability and availability analysis:
  • Qualitative analysis:
    • Hazard risk analysis
    • Failure mode and effects analysis (FMEA)
  • Reliability prediction:
    • Electronics FMEA
    • Fault tree analysis
  • Statistical simulations:
    • Maintainability analysis
    • Integrated logic support

Reliability engineering techniques can be used for reliability, availability, maintainability, and safety purposes. As this chapter is dedicated to reliability and availability engineering of data center ...