Abstract A distributed system can be characterized by the fact that the global state is distributed and that a common time base does not exist. However, the notion of time is an important concept in every day life of our decentralized \ r eal world" and helps to solve problems like getting a consistent population census or determining the potential causality between events. We argue that a linearly ordered structure of time is not (always) adequate for distributed systems and propose a generalized non-standard m o del of time which consists of vectors of clocks. These clock-vectors are p artially ordered a n d form a lattice. By using timestamps and a simple clock update mechanism the structure o f c ausality is represented in an isomorphic way. The new model of time has a close analogy to Minkowski's relativistic spacetime and leads among others to an interesting characterization of the global state problem. Finally, we present a new algorithm to compute a consistent global snapshot of a distributed system where messages may be r eceived out of order.

: Reasoning about knowledge seems to play a fundamental role in distributed systems. Indeed, such reasoning is a central part of the informal intuitive arguments used in the design of distributed protocols. Communication in a distributed system can be viewed as the act of transforming the system's state of knowledge. This paper presents a general framework for formalizing and reasoning about knowledge in distributed systems. We argue that states of knowledge of groups of processors are useful concepts for the design and analysis of distributed protocols. In particular, distributed knowledge corresponds to knowledge that is "distributed" among the members of the group, while common knowledge corresponds to a fact being "publicly known". The relationship between common knowledge and a variety of desirable actions in a distributed system is illustrated. Furthermore, it is shown that, formally speaking, in practical systems common knowledge cannot be attained. A number of weaker variants...

Agreement on the membership of a group of processes in a distributed system is a basic problem that arises in a wide range of applications. Such groups occur when a set of processes co-operate to perform some task, share memory, monitor one another, subdivide a computation, and so forth. In this paper we discuss the Group Membership Problem as it relates to failure detection in asynchronous, distributed systems. We present a rigorous, formal specification for group membership under this interpretation. We then present a solution for this problem that improves upon previous work.

The study of knowledge is of great use in distributed computer systems. It has led to better understanding of existing algorithms for such systems, as well as the development of new knowledge-based algorithm.~. The ability to achieve certain states of knowledge (e.g., common knowledge) provides a powerful tool for designing such algorithms. Un-fortunately, it has been shown that for many systems it is impossible to achieve these states of knowledge. In this paper we consider alternative interpretations of knowl-edge under which these states can be achieved. We explore the notion of consistent interpretations, and show how they can be used to circumvent the known impossibility results in a number of cases. This may lead to greater applicability of knowledge-based algorithms.

Communication concurrency is an application level property that is characterized completely by the set of messages exchanged in the application and the causal ordering relationship among the messages. The concurrency may be viewed at the logical level in the application rather than at the physical message communication level. The concurrency has a direct relationship to the message delivery performance in the underlying computation such as asynchronism in message delivery protocols. The paper quantitatively analyzes how the application level concurrency influences the execution of the application. The analysis is based on methods for representing concurrency and deriving measures of concurrency from the representations. System level execution models based on ISIS, x-Kernel and our causal broadcast communication system are used as case studies in the analysis. The analysis is in general useful in the design of applications and compare their projected performance levels in an implementat...

A distributed application runs as a collection of processes on a distributed system. This chapter presents a model of a distributed computation and introduces several terms, concepts, and notations that will be used in the subsequent chapters. 1

We propose a global flush communication prim-itive which allows the sender to order receipt of a message with respect to receipt of messages in its past and/or its Juture. The past of a message includes all messages that were sent /received by its sender before it is sent and all messages in their respective pasts. The Juture of a message includes all messages sent by its sender after it is sent or are sent by othrr processes after receiving it or messages sent after it, and messages in their respective futures. Use of this primitive provides a convenient way to reason about rnessagc orderings (as it allows a process to deduce the receipt event order-ings at other processes), simplifies algorithm de-sign and development, and permits higher com-munication level concurrency than communica-tion primitives proposed in the literature. We illustrate its applications and advantages, and present a counter based implementation.