Emulation of Software Faults: A Field Data Study and a Practical Approach
IEEE Transactions on Software Engineering
Information Assurance: Dependability and Security in Networked Systems
Information Assurance: Dependability and Security in Networked Systems
| Hi-index | 0.00 |
The most critical component of a dependabilitybenchmark is the faultload, as it should represent arepeatable, portable, representative, and generallyaccepted set of faults. These properties are essential toachieve the desired standardization level required by adependability benchmark but, unfortunately, are veryhard to achieve. This is particularly true for softwarefaults, which surely accounts for the fact that thisimportant class of faults has never been used in knowndependability benchmark proposals. This paper proposesa new methodology for the definition of faultloads basedon software faults for dependability benchmarking.Faultload properties such as repeatability, portabilityand scalability are also analyzed and validated throughexperimentation using a case study of dependabilitybenchmarking of web-servers. We concluded thatsoftware fault-based faultloads generated using ourmethodology are appropriate and useful for dependabilitybenchmarking. As our methodology is not tied to anyspecific software vendor or platform, it can be used togenerate faultloads for the evaluation of any softwareproduct such as OLTP systems.