Software Fault Tolerance Methods
Software fault tolerance methods initiate from fault tolerance designs in traditional hardware systems that require higher levels of dependability, reliability and availability.
In such systems, spare areas and backup units are generally used to keep the systems in operational conditions.
The primary software fault tolerance methods comprise N-version programming, recovery blocks and their variations.
There are some fundamental aspects of software fault tolerance methods:
- N-version programming uses parallel superfluity. In this case N copies, each of a different version of programs performing the identical functionality are running in parallel. The decision algorithm assures that local failures in limited quantity of such parallel versions will not compromise global implementation results.
- Recovery blocks use recurring implementations as the main mechanism for fault tolerance. If dynamic failures in some local parts are identified, a helping of the latest implementation is repeated, in the hope that this repeated implementation will not lead to the identical failure. Consequently, local failures will not spread to global failures.
In the majority of fault tolerance methods faults are not characteristically detected, consequently not removed, but just tolerated dynamically.
This is extremely different to software bugs identifying and removal activities such as inspection and software testing.
Learn more from QATestLab
Related Posts:
About Article Author
view more articles
No Comments Yet!
You can be the one to start a conversation.