Controlled Unreliable Process with Explicit or Implicit Breakdowns and Mixed Executive Times
This work continues the developments of the autors [5,10], concerning the minimization of the total executive time for unreliable processes until their correct finish by the help of suitable introduced controll schedule of tests, copies and check-points. It is supposed that there is an input flow of tasks (jobs, problems, service times) which must be executed for a given time X on a given apparatus (server, computer system etc.). The value X can be a mixture X = p 1 X 1+...+p r X r of r different values X 1,...,X r , where p i and X i are known, but it is never previously clear which value of X i will occur (we say the task is of type i, i=1,...,r). Simultaneously during the executive time X some undesirable events (breakdowns, catastrophes) can arize in a random way and lead to execution interruption or to incorrect final results. To have a guaranteed correct final result one needs a control test to detect the appearance of incorrectness in the case of implicit breakdowns. The execution must be repeated from the origin until no breakdown is detected. In the case of explicit breakdowns the repetitions from the origin are necessary until no breakdown happens during the execution time. These repetitions make the total executive process duration τ(X) eventually greater than X.
KeywordsPrincipal Minor Correct Finish Intal Blocking COpy Time Unreliable Process
Unable to display preview. Download preview PDF.
- 1.K.Barosov. An Optimal Control of the Service in a System with Evident and Latent Service Breackdown, Math. and athematwcalEducation,Sofia (1987), 315–319.Google Scholar
- 2.G.L.Brodezki. Effectivness of Storing Intermediate Results in Systems with Refasuals Distroying the Information, Izv. Acad. Sci. USSR, Technical Cibernetics 6, (1978), 97–103. (In Russian).Google Scholar
- 3.K.M.Chandy, J.C.Browe, C.W.Dissly, W.R.Ohrig. Analytical Models for Roll-back and Recovery Strategies in Data Base Systems, IEEE Trans. Software Eng. 1,(1975), 100–110.Google Scholar
- 5.B.N.Dimitrov and P.G.Petrov. Controlled Process with Explicit or Imlicit Breackdowns and Repeat Actions, In “Fault Diagnostic and Reliability”,Vol.1, Edit. Sp.Tzafestas at al., J.Reidal Publ.Comp., Dordrecht-Holland, (1987) (to appear).Google Scholar
- 6.DOS Supervisor and I/O Macros. No GC 24–5037–12. IBM System Reference Library.Google Scholar
- 7.V.I.lin, V.Sciduvui.Lchi, B.Sendov. Mathematical Analysis, v. 1, (1979), p. 744. (in Russian).Google Scholar
- 8.V.Hadjinov. Determination of optimal time intervals betwen check-points when computations are performed onspecialized multiprocessors, Electronic modelling 7, No 2 (1985), 14–18 (in Russian).Google Scholar
- 9.I.A.Kovalenco and L.S.Stoykova. On the System Productiveness and the Solution Time by Random Refusals and Periodical Storing of Results, Cybernetica 5, Kiev (1974), 73–75. (In Russian).Google Scholar