Metrics of System Dependability

Metrics To Achieve Dependability

There are four approaches to attain dependability, and are the following (Laprie, 1995):

• •

  Fault Prevention is an approach to enhancing development processes aiming to detect and reduce the number of faults in the system and eliminate the causes via process variations;

• •

  Fault Tolerance is when the system still provides the service to satisfy the system function in presence of faults;

• •

  Fault Removal is decreasing the presence of faults;

• •

  Fault Forecasting is an estimation of the current number, and the possible consequences of the faults.

Metrics of dependability are all the proprieties related to the system to be a dependable one. Again, a system is not dependable in proportion to the level of lowliness of these attributes. Moreover, and in the majority of cases, dependability is not a binary event but grounded on successions and acceptable approaches. These approaches are specific to infrastructures as well as applications. Among the dependability attributes, some need to be enhanced over others in specific system applications (Aqel, 2021).

Rams

Nowadays, engineers are confronting the ultimatum of high dependable system with the lowest cost of operations. This is where reliability, availability, maintainability and safety simultaneously come in. RAMS is an abbreviation for reliability, availability, maintainability, and safety which are defined as follows (Birolini, 2017):

• •

  Reliability is the ability of a functional unit (system) to perform a required function under given conditions for a given time interval (“ISO/IEC 2382-14:1997 Information technology Vocabulary Part 14:,” 1997). The reliability is expressed by the function R(t)=P(E) that E is a non-faulty entity on [0, t]. Supposing that E is not faulty at time t=0:R(t) = P(T>t)(1)