Fred B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):179--195, April 1982.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....3 A Naive Broadcast Based Replica Control Protocol We first develop a simple protocol for maintaining a replicated database in an idealized environment when there are no failures. Given that the underlying communication system supports totally ordered atomic broadcasts, the state machine approach [14] can be used to maintain replicated data. In the state machine approach operations are processed one at a time at every site in the same order. We can adopt this approach by requiring that a transaction initiated at a site broadcasts all its operations to all other sites using the atomic broadcast ....

F. B. Schneider. Synchronization in Distributed Programs. ACM Transactions on Programming Languages and Systems, 4(2):125--148, April 1982.

....areas, airport tower complexes to coordinate takeoffs and landings, and weather computer systems. As might be expected, the availability requirements for this system are 4 A message for which acknowledgmentshave been received is called a stable message in [PBS89] and fully acknowledged in [Sch82] stringent; for example, certain critical services are not supposed to be unavailable for more than 3 seconds per year. Such specifications have lead to the extensive use of fault tolerance techniques. The software associated with an ACCC is organized as a collection of services, each of which ....

F. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):125--148, Apr 1982.

....a non deterministicchoice construct. Silberschatz [22] and Van de Snepscheut [23] examined cases where the construct is easy to implement if only certain processes use it. Inefficient implementations of the general construct include those that use global information (e.g. a central coordinator) [19], require an unbounded amount of time [9] or use an unbounded amount of communication [2, 20] An overview of these techniques can be found in [4] Buckley and Silberschatz [4] were the first to propose a protocol that avoids all three of the inefficiencies above. Bagrodia [1] later proposed a ....

F. B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Language and Systems, 4(2):125--148, April 1982.

....work has also addressed mutual exclusion using more primitive read and write atomicity; the complexity of the resulting solutions is the principal motivation for the development of fetch and Phi primitives. Other researchers have considered mutual exclusion in the context of distributed systems [31, 39, 43, 45, 46], but the characteristics of message passing are different enough from shared memory operations that solutions do not transfer from one environment to the other. Our pseudo code notation is meant to be more or less self explanatory. We have used line breaks to terminate statements, and indentation ....

F. B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):179--195, Apr. 1982.

.... thereby simplifying the problems associated with failures [Cri91, Bir85a, EL90, KGR91, MPS92] A time service provides consistent information about time in a distributed system, either in the form of logical or virtual time that can be used to reason about the relative order of events [Lam78, Sch82, 25 Mat89] or real time from synchronized clocks [KO87, WL88, RSB90, VR92] Other important service abstractions are atomic actions, a collection of operations whose execution is indivisible despite concurrency and failures [Lam81, Lis85, SDP89] and stable storage, storage whose contents is ....

F. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):125--148, Apr 1982.

....[LJ94] presented a stepwise and modular method for the development of faulttolerant reactive systems: it showed how to design a component of a reactive system to tolerate both software design faults and the hardware faults of a given component. It also dealt with fault tolerant broadcasts [Sch82, SG84] and Byzantine agreement [LSP82] The earlier work did not deal with real time, but this could be added in a similar way to that shown here. Suppose that fault tolerance was to be achieved using checkpointing and backward recovery. As proposed in [LJ93] this can be done by a transformation of a ....

F.B. Schneider. Fault-tolerant broadcasts. ACM Transactions on Programming Languages and Systems, 4(2):125--148, April 1982.

....blocked will never be chosen. Each of these differences makes implementing events quite difficult. Reppy gives a coroutine implementation [31] but many of the difficulties arise in the presence of true parallelism. In particular, the last problem has been a topic of quite a bit of research [8, 34, 35]. Nevertheless, we have been able to implement events using the thread interface [27] The fact that so many higher level constructs can be efficiently implemented in terms of the thread interface (with help from SML s first class functions, polymorphic types, and module system) reinforces our ....

F. B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):125--148, April 1982.

No context found.

F. B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):179--195, April 1982.

No context found.

F. B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):179--195, Apr. 1982.

No context found.

Lainport, L. and F.B. Schneider. The 'Hoare Logic' of CSP, and All That. ACM Transactions on Programming Languages and Systems 6, 2 (April 1984), 281-296.

....detected any failure. Although this has an overhead (duplicated work when the EA or QE is not faulty but merely slow) it permits the system to ride out the period between when a failure occurs and when it is detected. Note that our approach does not fall into the category of active replication [14] either. Active replication means that the primary as well as all the backups act immediately and simultaneously. In our case, the backup does not act immediately, and whenever a reply is received before half the deadline, the backup does not reissue the request at all. 3.3 Achieving High ....

F. Schneider. Synchronization in Distributed Programs. ACM Transactions on Programming Languages and Systems, 4(2):125--148, 1982.

....detected any failure. Although this has an overhead (duplicated work when the EA or QE is not faulty but merely slow) it permits the system to ride out the period between when a failure occurs and when it is detected. Note that our approach does not fall into the category of active replication [9] either. Active replication means that the primary as well as all the backups act immediately and simultaneously. In our case, the backup does not act immediately, and whenever a reply is received before half the deadline, the backup does not reissue the request at all. 2.3 Achieving High ....

F. Schneider. Synchronization in Distributed Programs. ACM Transactions on Programming Languages and Systems, 4(2):125--148, 1982.

No context found.

Fred B. Schneider. Synchronization in distributed programs. ACM Transactions on Programming Languages and Systems, 4(2):179--195, April 1982.

No context found.

Fred B. Schneider. Synchronization in Distributed Programs. ACM Transactions on Programming Languages and Systems. 4:2 (April 1982), 179195.

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