Dependable Computing

7Measurement‐based Analysis of System Software: Operating System Failure Behavior

7.1 Introduction

The dependability of a computing system, and consequently of the services it provides to the end‐user, depends largely on the failure resilience of the underlying operating system. Understanding a system's sensitivity to errors and identifying its error propagation patterns and single points of failure are of primary importance in selecting a computing platform and assessing trade‐offs involving cost, reliability, and performance.

The dependability of a system can be experimentally evaluated at different phases of its lifecycle. In the design phase, computer‐aided design (CAD) environments are used to evaluate the design via simulation, including simulated fault injection. The feedback from simulation can be instrumental in the cost‐effective redesign of the system. Simulation, however, requires accurate input parameters and validation of output results. Although the parameter estimates can be obtained from past measurements, doing so is often complicated by design and technology changes.

In the prototype phase, the system runs under controlled workload conditions. In this stage, controlled physical fault injection is used to evaluate the system behavior under faults, including the detection coverage and the recovery capability of various fault tolerance mechanisms. Fault injection on the real system can provide information about the failure process from fault occurrence to system ...