Lehre.TestDep (Struktur)

Test und Dependability (3V1Ü)

(G. Kemnitz)

IT systems automate intellectual tasks: operational procedures control of processes and machines, design tasks, ... The basic requirement for use is dependability.
Key to this are checks and the and the elimination of detected deficiencies on three levels::
  • during design and manufacturing (fault prevention),
  • before use and during maintenance (troubleshooting), and
  • during operation (fault tolerance, damage prevention).
A key role here is played by the type and scope of the tests carried out and a test-oriented design as a prerequisite for ensuring that a system can be tested sufficiently. The content and learning objective of the lecture are the description of and measures for ensuring the dependability of systems consisting of hardware and software: monitoring, tests, problem elimination iterations, ... In practice, this is one of the most expensive and time-consuming tasks in the development and operation of IT systems.

Target audience of the lecture: Master of Computer Science and interested students from other degree programmes

Credits: (ECTS): 6

Slide sets [dt]

[S] -- Slide sets, [H] -- Handouts for printing.
  • [S1] Hazards, countermeasures and parameters
    1. Introduction
    2. Dependability: service model, availability, reliability, safety.
    3. Dealing with malfunctions: Monitoring, format checks, value checks, handling of detected malfunctions, dependability improvement.
    4. Fault elimination: Elimination iteration, Fault diagnosis and isolation, Test, stuck-at faults, reliability improvement, maturing processes, modular test, yield and defect level.
    5. Fault prevention: fault emergence, determinism and randomness, projects and process models, quality and creativity.
  • [S2] Probabilities
    1. Probabilities: definition and estimation, chained events, fault tree analysis, Markov chains.
    2. Fault detection: without and with memory, fault modelling.
    3. Troubleshooting: replacement, repair, maturing processes.
    4. Fault emergence.
  • [S3] Distributions
    1. Distribution: Characteristics, Linear transformation, distribution of counts
    2. Approximations for count distributions: Binomial distribution, Poisson distribution, Range estimation for Poisson distributed random variables, normal distribution, Range estimation for normally distributed random variables, variance increase, Range estimation with dependencies.
    3. Mixed distribution: properties, applications, range estimates with Chebyshev's inequality.
    4. Pareto distribution: properties, fault detection length, damage due to malfunctions.
    5. Gamma and exponential distribution
    6. Failures: Parameters, pre-ageing, reserve units, maintenance.
  • Tests and checks
    1. Test: inspection, function test, digital circuits, software, printed circuit boards.
    2. Monitoring: comparison, information redundancy, fault detecting codes, check indicators, protocols, time monitoring, invariants, syntax.
    3. Fault tolerance: error-correcting codes, RAID and backup, redundancy, system solutions.
  • Circuit test and self-test
    1. Fault modelling: circuit faults, local faults, fault models for digital circuits, detection relationships.
    2. Tests search: Fault simulation, D-algorithm, implication test, search space structuring, complex function blocks, sequential circuits, memory test.
    3. Self-test: pseudo-random register, signature register, self-test with LFSR, fault-oriented weighting, RAM self-test.
  • Software
    1. Programming language: data objects, control flow, malfunction handling, test.
    2. Design process: Software architecture, design flow, testable requirements, coding and testing..
    3. Test selection: Mutations, control flow, def-use chains, equivalence classes, UW graph, automata.
  • Exercises

References

News:

Old exams.


Autor: gkemnitz, Letzte Änderung: 15.04.2024 16:05:18

 TU Clausthal 2020  Impressum