Efficient communication of commitment-dependency information in the programmer-transparent coordination (PTC) scheme for cooperative recovery

Abstract

The Programmer-Transparen t Coordination (PTC) scheme is an approach to facilitating efficient cooperative backward recovery from errors caused by software faults and latent hardware faults in distributed computer systems (DCS's). The scheme allows independent and uncoordinated design of error detection and rollback capabilities of distributed processes. It makes provision for properly coordinating such distributed processes for cooperative recovery without incurring a domino effect. It does so by creating and maintaining a minimal number of recovery points. The coordination involves recording, inter-node communication, and updating of information on commitment-dependency that develops dynamically among the segments of distributed processes. The efficiency of handling the information on commitmentdependency is thus a major factor that determines the overhead incurred by the PTC scheme. The purpose of this paper is to establish a new principle on efficient identification and inter-node communication of the dynamically changing commitment-dependency among the distributed processes. Previously the principle of representing and communicating the commitmentdependency information in the form called the direct potential recaller set (DPRS) descriptor was established. It is shown in this paper that it is sufficient to maintain and communicate a part of the DPRS descriptor called the primary direct potential recaller set (pDPRS) descriptor. This could mean substantial reduction in both time and storage overhead associated with the PTC scheme in many applications.