S.K. Shrivastava, L.V. Mancini and B. Randell. "On the Duality of Fault-Tolerant System Structures," in Workshop on Experience with Distributed Systems (Lecture Notes in Computer Science, Vol. 309), pp. 19-37, Kaiserslautern, Springer-Verlag, 1987.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....a CA action is forbidden. On the other hand, exception resolution between exceptions raised in different roles might not be necessary from a strict transaction point of view and therefore the overhead in design and implementation complexity can be saved. Classification of Fault Tolerant Systems [21] partitions fault tolerant systems incorporating backward error recovery into two broad classes. The authors describe an object and action model, which is based on the use of transactions controlling operations on objects, and the process and conversation model, which consists of concurrent ....

Shrivastava, S. K.; Mancini, L. V.; Randell, B.: "The Duality of Fault-tolerant System Structures". SoftwarePractice and Experience 23(7), pp. 773 -- 798, July 1993.

....This is even more true when considering adding support for fault tolerance. Two different forms of atomic units have evolved: transactions and their derivatives which emphasize competitive concurrency, and conversations and their derivatives which emphasize cooperative concurrency. The authors of [5] name the former Object and Action model and the latter Process and Conversation model. They claim that the two models are duals of each other, and provide a mapping from one model to the other. Using this mapping, they show that mechanisms used in one model can have interesting counterparts in ....

Shrivastava, S. K.; Mancini, L. V.; Randell, B.: "The Duality of Fault-tolerant System Structures". Software -Practice and Experience 23(7), pp. 773 -- 798, July 1993.

....described in Chapter 1. Even though restartable actions are described in terms of a single process executing on one processor, the paradigm can be extended to systems of interacting processes in a distributed system. The global state of such a system is a set of states, one from each process [SMR88] This global state can be checkpointed by individually checkpointing each processes. However, the checkpointing and recovery of these processes must be coordinated with each other to avoid situations where the restart of a failed process p results in an inconsistent global system state and ....

S.K. Shrivastava, L.V. Mancini, and B. Randell. On the duality of fault tolerant system structures. In G. Goos and J. Hartmanis, editors, Experiences with Distributed Systems: Lecture Notes in Computer Science, Volume 309, pages 19--37. Springer-Verlag, Berlin, 1988.

....but their solution requires the duplication of the atomic actions of the application, the duplicate actions are used to obtain the global snapshots. In this paper we show algorithms for global snapshot that do not require such duplication. In this work we have used the duality of OAM and PCM [15] as a starting point for a detailed study of the precedence relations used to build consistent global snapshot algorithms in the OAM and in the PMM. The study of the duality OAM PCM motivated us to explore the correlations between control and communication mechanisms found in OAM and PMM. The ....

....properties of atomic actions (atomicity, consistency, isolation, and durability) provide a framework for building fault tolerant applications. These properties allow the integrity of objects to be maintained in the presence of failures such as node crashes fail silent nodes and message loss [15, 17]. The implementation of the atomic action mechanisms imply that objects must be controlled by an action manager m i that assigns read write locks to objects. We assume that atomic actions are organized according to action structures [16] determined by the application. The basic action structure ....

[Article contains additional citation context not shown here]

S. K. Shrivastava, L. V. Mancini, and B. Randell. The Duality of Fault-tolerant System Structures. Software---Practice and Experience, 23(7):773--798, July 1993.

....constructing fault tolerant distributed systems based on the concept of Coordinated Atomic (CA) actions [20] which provides uniform support for both cooperative and competitive concurrency. 1 Introduction The object and action model is used to provide fault tolerance object oriented systems [17]. Atomic transactions with the properties of serializability, failure atomicity and permanence of e ect are used to control operations on shared objects. The model attempts to hide the e ects of failures and concurrent processing from the application programmer, and is very e ective for certain ....

....on Zorzo s framework are analogous but each object is associated with a meta object. These meta objects implement several services such thread synchronization, exception resolution, design diversity and so on. 4 Transaction based approaches for cooperative concurrency The object and action model [17] is widely used to provide fault tolerance in distributedoriented systems. Several systems have been developed that successfully combine transaction processing with the object paradigm, e.g. Arjuna [16] These systems o er support for nested transactions and provide a powerful linguistic base for ....

S.K. Shrivastava, L.V. Mancini, and B. Randell. The Duality of Fault-Tolerant System Structures. Software - Practice and Experience, 23(7):773-798, July 1993.

.... (in which a sequence of changes on a set of data items are undone together) which allow the designer to manage together error recovery and concurrency control in accessing data [Lynch, Merrit et al. 1993] Conversations and atomic transactions are in fact dual models of recovery, as discussed in [Shrivastava, Mancini et al. 1993]: they are two ways of describing the same backward recovery philosophy in the two models (or design styles) which the authors of [Shrivastava, Mancini et al. 1993] call the object action model (where the long term state of the computation is encapsulated in data objects, and active processes ....

.... accessing data [Lynch, Merrit et al. 1993] Conversations and atomic transactions are in fact dual models of recovery, as discussed in [Shrivastava, Mancini et al. 1993] they are two ways of describing the same backward recovery philosophy in the two models (or design styles) which the authors of [Shrivastava, Mancini et al. 1993] call the object action model (where the long term state of the computation is encapsulated in data objects, and active processes invoke operations on these objects) and the process conversation model (where the state is contained in the processes, which communicate via messages) ....

[Article contains additional citation context not shown here]

S.K. Shrivastava, L.V. Mancini and B. Randell, "The Duality of Fault-Tolerant System Structures", Software Practice and Experience, vol. 23, no. 7, pp.773-798, 1993.

....a means for isolating entities in a system characterized by competitive activity) Thus, the conversation has to be a software engineering facility and not just a facility for describing recovery. That is why we believe that conversations should be the units of software structuring. The authors of [3] discuss the duality of the conversation of processes which can also be treated as a set of interconnected objects involved in a transaction (in object based systems) We believe that the next step should be made to give programmers a facility to implement concurrent OO systems with diversity. A ....

Shrivastava, S. K., Mancini, L. V. and Randell, B. The Duality of Fault-Tolerant System Structures. Software Practice and Experience 23: 773-798; 1993.

....of computation from different objects are designed together. They are supposed to be executed concurrently and to use the same servers (for information exchange and synchronization) To regard a conversation in an object based system as a set of interconnected objects is implicitly proposed in [19], but there are no specific proposals for implementation, the computational model or the language constructions for conversations. We believe that the right way of understanding the problem of conversations in COOLs is to treat objects as conversation participants and as a communication tool. This ....

Shrivastava, S. K., Mancini, L. V. and Randell, B. The Duality of Fault-Tolerant System Structures. Softw. - Pract. Exper. 23: 773-798; 1993.

....it with the least possible hassle for the programmer. These techniques include the distribution of the application processes (problems involved in distribution were focused in [1] and [2] a journaling mechanism allowing the final user to do arbitrary replays of his session; fault tolerance (see [3], for instance) in case of communication failures of the network or some of the processes; a powerful control language allowing the developer to describe his application and aspects of its user interface and allowing the final user to introduce his own macros; and mechanisms for parametric and ....

Shrivastava, Mancini, and Randell. The Duality of Fault-Tolerant System Structures. Software Practice and Experience, 23(7), July 1993.

....processes to successively earlier checkpoints to find a consistent state. In addition to checkpointing facilities, some sort of reliable interprocess communication is also required to implement this model. The object action model and conversation model have been shown to be duals of one another [SMR88] Abstractions that are useful for supporting conversations include stable storage and resilient processes. An application following the primary backup approach is organized as a collection of services, each of which is implemented by multiple processes to provide fault tolerance. The name comes ....

S. K. Shrivastava, L. V. Mancini, and B. Randell. On the duality of fault tolerant system structures. In J. Nehmer, editor, Experiences with Distributed Systems, volume 309. LNCS Springer-Verlag, 1988.

....E. Buzato their solution requires the duplication of the atomic actions of the application, the duplicate actions are used to obtain the global snapshots. In this paper we show algorithms for global snapshot that do not require such duplication. In this work we have used the duality of OAM and PCM [15] as a starting point for a detailed study of the precedence relations used to build consistent global snapshot algorithms in the OAM and in the PMM. The study of the duality OAM PCM motivated us to explore the correlations between control and communication mechanisms found in OAM and PMM. The ....

....properties of atomic actions (atomicity, consistency, isolation, and durability) provide a framework for building fault tolerant applications. These properties allow the integrity of objects to be maintained in the presence of failures such as node crashes fail silent nodes and message loss [15, 17]. The implementation of the atomic action mechanisms imply that objects must be controlled by an action manager m i that assigns read write locks to objects. We assume that atomic actions are organized according to action structures [16] determined by the application. The basic action structure is ....

[Article contains additional citation context not shown here]

S. K. Shrivastava, L. V. Mancini, and B. Randell. The Duality of Fault-tolerant System Structures. Software --- Practice and Experience, 23(7):773--798, July 1993.

No context found.

S.K. Shrivastava, L.V. Mancini and B. Randell. "On the Duality of Fault-Tolerant System Structures," in Workshop on Experience with Distributed Systems (Lecture Notes in Computer Science, Vol. 309), pp. 19-37, Kaiserslautern, Springer-Verlag, 1987.

No context found.

S. K. Shrivastava, L. V. Mancini and B. Randell, `On the duality of fault-tolerant system structures', in J. Nehmer (ed.), Lecture Notes in Computer Science, Vol. 309, Springer Verlag, 1987, pp. 19--37.

....concurrent processes like the conversation scheme in order to be effective in actual systems, such coordinated actions need in turn to treat implicit interactions deliberately, especially between shared resources. Many dual aspects of conversations and atomic transactions are discussed in [30] together with some differences FTCS 25 Submission 17 between the two mechanisms. In [32] a proposal is made for dividing a heterogeneous system into two subsystems using conversations and transactions respectively. Within the context of OO languages and systems, our work instead offers a means ....

S.K. Shrivastava, L.V. Mancini and B. Randell, "The Duality of Fault-Tolerant System Structures," Software - Practice and Experience, vol.23, no.7, pp.773-798, 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC