Jalote, P. Using broadcast for multiprocess recovery. Proc. IDCSS-6, Massachusetts, 1986  Home/Search   Document Not in Database   Summary   Related Articles   Check

....for different applications, computational models, languages, failure assumptions, etc. In this paper we will concentrate on distributed systems. There are several distributed atomic action schemes, in all of which action participants are distributed and cooperate by message send receive. Paper [3] discusses how broadcasting can be used to implement a conversion scheme (BER is used here) it is assumed that the underlying system provides two operations: broadcast and receive. The concept of distributed real time conversations is introduced in [4] This approach uses software diversity: it ....

Jalote, P. Using broadcast for multiprocess recovery. Proc. IDCSS-6, Massachusetts, 1986

....in both research and practice to achieve this goal. Most of the existing schemes for exception handling in concurrent systems use the concept of an atomic action as a unit of error confinement, though there is no clear consensus on how to handle exceptions when asynchronous activities occur [Jalote Campbell 1986][Taylor 1986] Many new architectural developments in the area of distributed computing systems are, to some extent, object based or object oriented (OO) The OO technique, with its modularity, flexibility and reusability features, can be usefully exploited for handling complexity and ....

P. Jalote, "Using Broadcast for Multiprocess Recovery," In Proc. 6th Distributed Computing Systems Symposium, pp.582-589, 1986.

....multiple exceptions raised concurrently and to signal exceptions over nested actions. 6.1 Related Work There has been relatively little work on implementations of coordinated error recovery in a distributed system. Implementations of distributed process oriented conversations are discussed in [Jalote 1986][Yang Kim 1992] The Arche language introduced in [Issarny 1993] allows the programmer to implement a function that can resolve the exceptions propagated from several objects of the same type. Such resolution is however only suitable for a limited form of concurrency. Wellings and Burns ....

P. Jalote, "Using Broadcast for Multiprocess Recovery," In Proc. 6th Distributed Computing Systems Symposium, pp.582-589, 1986.

....(a) X handling P1 P2 P3 (b) aborting exception exception Figure 1. Two methods for treating nested actions while an exception raised. There has been relatively little work on implementations of distributed coordinated error recovery. Implementations of distributed conversations are discussed in [12][25] Of these [25] focused on two particular conversation schemes (i.e. the name linked recovery block and the abstract data type) The work in [12] discussed a distributed implementation of the conversation scheme using broadcasts, assuming that all processes enter a conversation simultaneously. ....

....relatively little work on implementations of distributed coordinated error recovery. Implementations of distributed conversations are discussed in [12] 25] Of these [25] focused on two particular conversation schemes (i.e. the name linked recovery block and the abstract data type) The work in [12] discussed a distributed implementation of the conversation scheme using broadcasts, assuming that all processes enter a conversation simultaneously. Both of these approaches cannot be used directly to implement the CA action scheme because they are directed into some particular schemes, with no ....

P. Jalote, "Using Broadcast for Multiprocess Recovery," In Proc. 6th Distributed Computing Systems Symposium, pp.582-589, 1986.

....exception exception Figure 1: Two methods for treating nested actions while an exception is raised There has been relatively little work on implementations of coordinated error recovery in a distributed system. Implementations of distributed process oriented conversations are discussed in [13][26] Of these [26] focused on two particular conversation schemes (i.e. the name linked recovery block and the abstract data type) and addressed various implementation issues which are specific to the chosen conversation structures. The work in [13] discussed a distributed implementation of the ....

....process oriented conversations are discussed in [13] 26] Of these [26] focused on two particular conversation schemes (i.e. the name linked recovery block and the abstract data type) and addressed various implementation issues which are specific to the chosen conversation structures. The work in [13] discussed a distributed implementation of the conversation scheme using broadcasts, assuming that all processes will enter each conversation simultaneously. Neither approaches can be used directly to implement the CA action scheme because they focus on some particular schemes, with no support for ....

P. Jalote, "Using Broadcast for Multiprocess Recovery," In Proc. 6th Distributed Computing Systems Symposium, pp.582-589, 1986.

....this scheme cannot be treated, form our point of view, as a distributed scheme. A very important drawback of the scheme is that connecting processes in a chain requires that processes enter the action in the order in which they are going to participate. An alternative scheme was also proposed by Jalote [1986]. This eliminates the process ordering and uses a central marked process through the use of multicast. When processes enter and leave the action (synchronously) they multicast entry and exit messages to all other participants. The participants enter the action synchronously. This scheme is the ....

P. Jalote, "Using broadcast for multiprocess recovery," presented at 6th Distributed Computing Systems Symposium, Cambridge, Massachusetts, 1986.

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