Skip to main content

Research Repository

Advanced Search

Process survivability in a distributed computer control system

Trueman, Peter

Process survivability in a distributed computer control system Thumbnail


Peter Trueman


Possibly the greatest advantage that a distributed computer control system has over a centralised control system is that the failure of one or more of its constituent computers does not prevent the other computers from operating normally. Unfortunately, the loss of the executive and application software hosted by a failed computer will prevent the surviving part of the control system from fulfilling its role. Whereas it is possible to design the executive software so that the loss of one of its constituent kernels will not prevent the others from functioning normally, it is not possible to do this for the application software.
Process survivability was conceived as a way of preventing application processes from being lost as a result of a computer failure. Process survivability enhances the high availability of a distributed computer control system’s hardware by making its application software invulnerable to computer failure. Process survivability is performed in a way that is transparent to the application programmer.
In this thesis we first describe the distributed computer control system called PROSUP (PROcess SURvivability) which we designed as an environment in which to develop process survivability.
The major part of this thesis is concerned with the design and development of process survivability for PROSUR. In particular, we describe how redmdant inactive copies of all of the application processes are incorporated into the application software and how the processes are recovered to a consistent state after a computer failure.
As well as showing that process survivability is practicable, we also investigate its practicality. A simulation study of a distributed computer control system incorporating process survivability has been performed to gain an insight into the effects that process survivability might have on a control system's performance. The results of this simulation are presented and a number of interesting conclusions are drawn.


Trueman, P. (1985). Process survivability in a distributed computer control system


Downloadable Citations