Approaches to design of temporary blackout handling capabilities and an evaluation with a real-time tightly coupled network testbed

• Published in: [1991] Digest of Papers. Fault-Tolerant Computing: The Twenty-First International Symposium pp. 470 - 477
• Main Authors: Kim, K.H., Guan, W.J., Damm, A., Rohr, J.A.
• Format: Conference Proceeding
• Language: English
• Published: IEEE Comput. Soc. Press 1991
• Subjects: Application software; Distributed computing; Distributed control; Electronic components; Local area networks; Read-write memory; Real time systems; Registers; Software design; Testing
• ISBN: 9780818621505; 0818621508
• DOI: 10.1109/FTCS.1991.146703

The problem of designing real-time tightly coupled networks (TCNs) that can survive through temporary blackout (TB) events that disrupt orderly operation of electronic components and erase the contents of all registers and RAMs is treated. Three approaches are considered. The first, in existence for some time although not practiced extensively, relies on the software designer for highly application-dependent design of all parts of TB handling procedures. The second is a fully application-transparent approach to state saving based on the use of recoverable regions. The third, a compromise between the first two approaches, uses time-driven insertion of saving points and relies on the software designer to determine the membership of the state vector. An experiment that involved designing TB handling capabilities into a real-time TCN testbed using each approach is reported. The testbed-based evaluation results validated the practicality of the two new approaches and indicated the complementary relationship between them and the first approach.< >