A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Workshop on Distributed Algorithms, pages 219-- 232, Nice, France, 1989.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....This results in simplicity in design of distributed algorithms which may depend on the FIFO assumption [3] In this paper, we discuss one such possible restriction on message ordering called synchronous message ordering. The problem of restricting message ordering has received wide attention [1, 7, 12, 13, 10]. Charron Bost, Mattern, and Tel [1] have recently proposed a strict hierarchy of message ordering asynchronous, FIFO, causal ordering, and synchronous. An asynchronous computation does not have any restriction on the message ordering. It is easy to implement, and it permits maximum ....

....that a single message should not be overtaken by a sequence of messages. Joseph and Birman [7] have given many examples of problems which are easier to solve if causal ordering is assumed. It was first implemented in ISIS [4] Many other algorithms for causal ordering have appeared since then [12, 13]. Synchronous ordering is a stronger requirement than causal ordering. Algorithms for synchronous systems are much easier to design than those for causally ordered systems. Unlike the case of causal ordering, there is very little work done on synchronous ordering of messages. Usually, to achieve ....

Schiper, A., Eggli, J., and Sandoz, A., "A New Algorithm to Implement Causal Ordering", Proc. of the 3rd Int. Workshop on Distributed Algorithms, pp. 219-232, Springer Verlag, 1989. 19

.... such as management of replicated data [8, 9] distributed monitoring [6] resource allocation [18] distributed shared memory [3] multimedia systems [2] and collaborative work [19] The protocols to implement causal message ordering in systems with static hosts have been presented in [14, 9, 16, 18, 20, 21]. These protocols can be executed by every mobile host with all the relevant data structures being stored on the mobile hosts themselves. However, considering limited resources and bandwidth of wireless links available to mobile hosts, it is not appropriate to apply these protocols directly to ....

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Proceedings of the 3rd International Workshop on Distributed Algorithms, LNCS-392, pages 219-232, Berlin, 1989.

....m1 :t m2 :t. However, m1 and m2 can be delivered in any order if m1 :t k m2 :t. Causal message ordering has been an active research topic in many areas such as distributed databases, realtime systems, and fault tolerant systems. Examples include solutions for distributed systems in general [15, 25], for mobile computing systems [22, 27] and as part of total ordering multicast protocols with support for causal order delivery of messages from multiple sources [2, 7, 10, 20] However, causal message ordering is not sufficient in wireless sensor networks, since it is based on logical time. 4. ....

....(x; y; z) of the sensor node, and optionally additional data such as sensor readings. Producers advertise sensor events by specifying the names and possible locations of sensor events they are going to generate. Consider the following two advertisements: advertise detected at [10,15] 0,10][20,25]; advertise lost at [10,15] 0,10] 20,25] which announce that a producer will publish detected and lost events with 10 x 15, 0 y 10, and 20 z 25, which represents the coverage area of its sensors. The two events might represent detection a tracked object ( detected ) and no ....

[Article contains additional citation context not shown here]

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Workshop on Distributed Algorithms, pages 219--232, Nice, France, 1989.

....for a class of message ordering specifications. This is the focus of the companion paper [19] 2 Related Work A fair amount of research has been done for efficient algorithms to implement different message orderings. Birman and Joseph [4] Raynal, Schiper and Toueg [20] Schiper, Eggli and Sandoz [21], have presented algorithms for the causal ordering of messages. These algorithms tag knowledge of processes about messages sent in the system with the message. For example, process P i in the algorithm by Raynal, Schiper and Toueg [20] tags a message with the matrix m where m[j; k] is the ....

...., the set of all partial orders as X async = f (H; x:s 2 H , x:r 2 H) and . is a partial order g : Causal Ordering (CO) Causal ordering can be stated as s 1 . s 2 ) r 2 . r 1 ) There exists a tagged algorithm where with each message a matrix of size n Theta n is tagged to the message [20, 21]. Formally, we can state X co , the set of partial orders satisfying causal ordering as X co = f (H; x:s . y:s) y:r . x:r) 8 x; y 2 M g : Logically Synchronous (SYNC) A run is logically synchronous if its time diagram can be drawn such that all message arrows are vertical. ....

[Article contains additional citation context not shown here]

A. Schiper, J. Eggli, and A. Sandoz. A new algorithm to implement causal ordering. In Proceedings of the Third International Workshop on Distributed Algorithms, pages 219--232. Springer-Verlay, 1989.

....are caused by message exchanges between event generating processes. In sensor networks, event notifications are generated based on real world events. The causal relationships (i.e. real world messages ) between these events cannot be captured by the WSN. Therefore, causal delivery mechanisms [13, 24] based on logical time are not applicable for our purposes in WSN. For the same reasons, previous work on causal order delivery in mobile computing systems [20, 28] is not applicable in settings, where we are interested in temporal relationships between real world events. The ideas for group ....

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Workshop on Distributed Algorithms, pages 219--232, Nice, France, 1989.

....messages meant for that destination have either been delivered or their deadlines have expired. This is the basic idea of the # causal ordering protocol proposed by Baldoni, Mostefaoui and Raynal [3] However, 3] and all general purpose causal ordering algorithms prior to that, including [1, 5, 14, 16], require each message to carry an # # # matrix that contains information about all causal predecessors of the message. In the tele immersive application described earlier most of the time a message will be sent to only a small subset of participants that are within a short distance of the ....

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm To Implement Causal Ordering. In Proceedings of the # ## International Workshop on Distributed Algorithms, LNCS-392, pages 219--232, Berlin, 1989. Springer.

....by capturing the causality relation between events in a distributed computation. VC, independently proposed by Fidge [12] and Mattern [20] is extensively used in distributed applications, such as, distributed debugging [13] checkpointing and recovery [16, 18] and causal communication [27]. Applications that use VC may require unbounded space, since vector clocks grow unboundedly. In order to implement VC applications using bounded space, VC should also be bounded, which leads us to make two observations about VC. First, VC often consumes more space than is necessary by requiring ....

A. Schiper, J. Eggli, and A. Sandoz. A new algorithm to implement causal ordering. Proceedings of the International Workshop on Distributed Algorithms, pages 219-232, 1989.

....ensures that if send p (m 1 ) Gamma p send q (m 2 ) then deliver r (m 1 ) Gamma p deliver r (m 2 ) i.e. m 1 is delivered to r before m 2 . 2 Well known causal delivery protocols are the causal broadcast multicast protocols using: piggybacking [2] context graphs [12] or vector clocks [3, 14]. These protocols, adequate for asynchronous systems, deliver messages according to a logical ordering [2] a message m 1 is said to logically precede ( l ) m 2 if m 1 is sent before m 2 , by the same participant or m 1 is delivered to the sender of m 2 before it sends m 2 or there exists a ....

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Proceedings of the 3rd Int Workshop on Distributed Algorithms, volume LNCS 392, pages 219--232, Nice - France, September 1989. Springer Verlag.

.... such as management of replicated data [8, 9] distributed monitoring [6] resource allocation [18] distributed shared memory [3] multimedia systems [2] and collaborative work [19] The protocols to implement causal message ordering in systems with static hosts have been presented in [14, 9, 16, 18, 20, 21]. These protocols can be executed by every mobile host with all the relevant data structures being stored on the mobile hosts themselves. However, considering limited resources and bandwidth of wireless links available to mobile hosts, it is not appropriate to apply these protocols directly to ....

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Proceedings of the 3rd International Workshop on Distributed Algorithms, LNCS-392, pages 219--232, Berlin, 1989.

.... such as management of replicated data [5, 6] distributed monitoring [4] resource allocation [11] distributed shared memory [2] multimedia systems [1] and collaborative work [12] The protocols to implement causal message ordering in systems with static hosts have been presented in [7, 6, 9, 11, 13, 14]. These protocols can be executed by every mobile host with all the relevant data structures being stored on the mobile hosts themselves. However, considering limited resources and bandwidth of wireless links available to mobile hosts, it is not appropriate to apply these protocols directly to ....

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Proceedings of the 3rd International Workshop on Distributed Algorithms, LNCS392, pages 219--232, Berlin, 1989.

....can be huge. The CO algorithm in [8] is similar to [2] but carries message ids rather than entire messages in the control information. Furthermore, unnecessary control information is not sent if the sending host had sent it before. The control information in the Schiper Eggli Sandoz CO algorithm [10] consists of n vectors of 1 The arrival of a message signifies that the communication network has placed the message in the buffer of the receiving process; Its delivery means that the process has taken up the message for processing 1 length upto n each. This information represents messages ....

....n with every message. SENT [i; j] indicates the number of messages that are known to be sent by i to j. The algorithm also uses an array DELIV of size n, where DELIV [i] is the number of messages from node i that have been delivered to the sender. Clearly, the message overhead of both algorithms [9, 10] is O(n 2 ) In the causal multicast in overlapping groups implementation of ISIS [3] every process maintains a vector for every group whether it belongs to that group or not. A vector for a group informs the process of the number of messages multicast by the various members of the group. When ....

A. Schiper, J. Eggli, A. Sandoz, A New Algorithm to Implement Causal Ordering, Proc. 3rd Int. Workshop on Distrib. Algorithms, 1989, 219-232, in LNCS 392, Springer-Verlag. 10

....is one such system [ISIS91] Causal ordering of messages is thought by many to greatly reduce the complexity, development time, and the debugging effort needed to implement distributed programs. The partial global ordering of messages according to causality is accomplished with logical clocks [Lam78, Mar84, SES89, BSS90]. Several types of logical clocks are also described in Chapter 2. 1 A more formal definition of causal ordering is given in Chapter Two. 4 1.2.2 Performance evaluat ion In performance evaluation, analysts are generally concerned with the passage of time, or with timing of intervals. Some ....

....are delivered in causal order. Causal ordering means that if event E 1 on processor P 1 causes an event E 2 on processor P 2 , then all processors in the system are guaranteed to see the effects (resultant messages) of events E 1 and E 2 in the proper order. The precise definition, taken from [SES89] follows. Define the relation happened before , denoted as , to be the transitive closure of the relation R. R is defined as: Two events, E 1 and E 2 , are related by R, iff any of the following two conditions are true: 1) E 1 and E 2 are two events occurring in the same process (E 1 before E ....

Schiper, A., Eggli, J. and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Proceedings of the 3rd International Workshop on Distributed Algorithms, Lecture Notes on Computer Science 392, pages 219-232. SpringerVerlag, 1989.

....are not applicable to as wide a range of topologies. 10 The causally ordered multicast is a generalization of causal message delivery to multicasts that has been found to be useful in distributed MBM programming applications [4, 26] Existing causal multicast protocols for non bus based systems [3, 4, 22, 31] require multiple message rounds. Isotach networks support a single round multicast [34] # Totally Ordered Multicasts A totally ordered multicast is a multicast that is received in a consistent order at all processes in the system. Total ordering is useful in MBM computations for a variety ....

A. Schiper, J. Eggli and A. Sandoz, "A New Algorithm to Implement Causal Ordering", in Distributed Computing, vol. 89 , Springer-Verlag, Berlin-Heidelburg-New York, 1989, 219-232.

....Group Communication in a Distributed System. be below the transport layer depending on the implementation. Some services provided by the group communication layer are described below. 1. Causal Ordering of Messages Causal ordering of messages invocations is a natural extension of FIFO ordering [13, 14, 15, 16, 17, 18, 19, 20]. It provides synchronization guarantees that make co operative distributed programming easier. A causal order is a partial order which follows directly from Lamport s Happened Before relation [21] Informally, the requirements of causal ordering are: If the send of a message m1 Happened Before ....

....In case of a network partition and a simultaneous primary failure, the algorithm does not rejoin the partitions correctly. Causal ordering for a fault tolerant system was first proposed in [35] Most Causal ordering algorithms use logical vector clocks proposed by Mattern and Fidge [36, 37] [14] and [13] propose causal ordering algorithms for non fault tolerant cases, but put a bound on the storage requirement. Kopetz et al. 38] have developed a protocol for the ordering of messages in a synchronous system for real time applications. They also provide a membership service for a ....

[Article contains additional citation context not shown here]

Schiper, Eggli, and Sandoz, "A new algorithm to implement causal ordering," in WDAG: International Workshop on Distributed Algorithms, Springer-Verlag, Berlin, 1989.

....need to have their cause and effect relations preserved. A causal order algorithm is required to preserve the causal order of the messages. Multicasting without causal ordering and real time constraints has been studied in ATM networks [5, 20] and causal order without multicasting has been studied [18, 21]. Our objective is to establish boundaries to determine, given a multicast network topology, which causal ordering method is most efficient. To that end, we focus on two algorithms with different properties, such as overhead, and type of causal ordering provided. In addition, we propose a new ....

....overheads. The computational overhead comes from the underlying algorithm that enforces causal order and varies from algorithm to algorithm. The message overhead comes from the extra control information that must be transmitted with each message. This information can be in the form of time vectors [18, 21] or message histories [2] ISIS, the first system to support causal multicasting, sends complete message histories with each message, which may cause a significant overhead. It enforces causal order of messages and additionally 9 can enforce a total order on all messages, including concurrent ....

[Article contains additional citation context not shown here]

A. Schiper, J. Eggli, A. Sandoz, "A New Algorithm to Implement Causal Ordering", In Proceedings of the International Workshop on Distributed Algorithms, 1989, in Lecture Notes in Computer Science 392, Springer-Verlag, New York, pp. 219-232.

.... reception, called causal ordering, is required for a variety of applications like management of replicated data, observation of a distributed system, resource allocation, multimedia systems, and teleconferencing [2, 4, 15] Protocols to implement causal ordering of messages have been presented in [5, 6, 14, 15, 17]. These protocols have high communication overheads. For each of these protocols (except [5] which is based on message duplication a high communication overhead approach) the message overhead is at least Theta(N 2 ) integers, where N is the number of processes in the system. Hence, the ....

....M is delivered to the destination. Garbage collection is required from time to time to eliminate the old messages that no longer need to be carried by the new messages. Otherwise, the number of predecessor messages carried by each message will grow indefinitely. The implementation presented in [17] uses vector clocks [9, 12] Unlike the first version of ISIS, each message M carries control information consisting of ordered pairs of the type (destination site, vector time) There can be up to N Gamma 1 such ordered pairs with each message. When a destination process receives M , it uses the ....

[Article contains additional citation context not shown here]

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm To Implement Causal Ordering. In Proceedings of the 3 rd International Workshop on Distributed Algorithms, LNCS-392, pages 219--232, Berlin, 1989. Springer.

....returns true if, for all i, V T 0 [i] V T 00 [i] and there is at least one component of V T 0 that is less than the corresponding component of V T 00 . Vector clocks have been used extensively in distributed system to do distributed debugging [22] achieving causal ordering of messages [53], building highly available distributed services [43] and checkpointing for optimistic recovery [32] among other applications. 5.1.2 The Basic Algorithm Attempts to read a location not locally owned nor cached (a read miss) generate a message to the owner requesting a current copy. The ....

A. Schiper, J. Eggli, and A. Sandoz. A new algorithm to implement causal ordering. In 3rd International Workshop on Distributed Algorithms, 1988.

....RAM) consistency [25] is a consistency criterion weaker than causal consistency. The difference, in the shared memory model, between PRAM consistency and causal consistency is the same as the one between FIFO ordering and causal ordering for message deliveries in the message passing model [14, 33, 31]. PRAM and FIFO are only concerned by direct relations between pairs of adjacents processes and do not take into account transitivity due to intermediary processes. More precisely, in a message passing system with FIFO ordering, two messages sent to a same process by two distinct senders can ....

....receive events to be ordered as their associated send events. It is easy to see that, if b H has causally ordered communications, it has also FIFO communications. Figure 5 illustrates causal ordering: in Figure 5.a communications are causally ordered, while they are not in Figure 5. b referencws [14, 33, 31] describe protocols implementing causally ordered communications on top of an asynchonous sytem, and give examples of problems whose solutions are easier to design when the underlying network ensures communications are causally ordered at the application level. 8.2.2 Logically Instantaneous ....

A. Schiper, J. Eggli, and A. Sandoz. A new algorithm to implement causal ordering. In Proc. 3rd Intl. Workshop on Distributed Algorithms (WDAG-3), Springer Verlag LNCS 392 (J.C. Bermond and M. Raynal Eds), pages 219-232, 1989.

No context found.

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Workshop on Distributed Algorithms, pages 219-- 232, Nice, France, 1989.

No context found.

A. Schiper, J. Eggli, and A. Sandoz. A New Algorithm to Implement Causal Ordering. In Workshop on Distributed Algorithms, pages 219--232, Nice, France, 1989.

No context found.

A. Schiper, A. Eggli, and A. Sandoz, "A New Algorithm to Implement Causal Ordering," Proc. Third Int'l Workshop Distributed Systems, pp. 219-232, 1989.

No context found.

A. Schiper, J. Eggli, and A. Sandoz. A new algorithm to implement causal ordering. In Proceedings of the 3rd International Workshop on Distributed Algorithms, 1989. Also published in Lecture Notes in Computer Science, 392.

No context found.

A. Schiper, J. Eggli, and A. Sandoz. A new algorithm to implement causal ordering. In J.-C. Bermond and M. Raynal, editors, Proceedings of the Third International Workshop on Distributed Algorithms, volume 392 of Lecture Notes on Computer Science, pages 219--232, Nice, France, September 1989. Springer-Verlag.

No context found.

Schiper, A., Eggli, J., and Sandoz, A. A new algorithm to implement causal ordering. Proc. Third International Workshop on Distributed Algorithms, 1989.

No context found.

SCES89 Schiper A., Eggli J. and Sandoz A., A New Algorithm to Implement Causal Ordering, Proceedings of the 3 rd International Workshop on Distributed Algorithms, Berlin, 1989, pp 219-232.

First 50 documents  Next 50

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