Results 11  20
of
22
Room Synchronizations
, 2001
"... We present a class of synchronization called room synchronizations and show how this class can be used to implement asynchronous parallel queues and stacks with constant time access (assuming a fetchandadd operation). The room synchronization problem involves supporting a set of m mutually exclusi ..."
Abstract

Cited by 7 (3 self)
 Add to MetaCart
We present a class of synchronization called room synchronizations and show how this class can be used to implement asynchronous parallel queues and stacks with constant time access (assuming a fetchandadd operation). The room synchronization problem involves supporting a set of m mutually exclusive "rooms" where any number of users can execute code simultaneously in any one of the rooms, but no two users can simultaneously execute code in separate rooms. Users asynchronously request permission to enter specified rooms, and neither the arrival time nor the arrival order nor the desired room of such requests are known ahead of time. We describe an algorithm for room synchronizations, and prove it satisfies a number of desirable properties. We have implemented our algorithm on a Sun UltraEnterprise 10000 multiprocessor. We present experimental results comparing an implementation of a parallel stack using room synchronizations to one using locks, demonstrating a significant scalability advantage for room synchronizations.
Bakery Algorithms
 Proceedings of the CS& P'93 workshop. pp.740, Warszawa
, 2001
"... An approach to proving higher level properties of distributed protocols is suggested here in which a proof consists of two stages: In the higherlevel stage, abstract properties of system executions are assumed and their desired consequences are proved. At the lowerlevel stage these abstract pr ..."
Abstract

Cited by 5 (4 self)
 Add to MetaCart
An approach to proving higher level properties of distributed protocols is suggested here in which a proof consists of two stages: In the higherlevel stage, abstract properties of system executions are assumed and their desired consequences are proved. At the lowerlevel stage these abstract properties are shown to hold in every execution of the protocol.
Computing with infinitely many processes under assumptions on concurrency and participation
 In 14th Int. Symp. on DIStributed Comp. (DISC
, 2000
"... We explore four classic problems in concurrent computing (election, mutual exclusion, consensus, and naming) when the number of processes which may participate is infinite. Partial information about the number of actually participating processes and the concurrency level is shown to affect the possi ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
We explore four classic problems in concurrent computing (election, mutual exclusion, consensus, and naming) when the number of processes which may participate is infinite. Partial information about the number of actually participating processes and the concurrency level is shown to affect the possibility and complexity of solving these problems. We survey and generalize work carried out in models with finite bounds on the number of processes, and prove several new results. These include improved bounds for election when participation is required (even for finitely many processes, as investigated by Styer and Peterson [SP89]) and a new adaptive starvationfree mutual exclusion algorithm for unbounded concurrency. We survey results in models with shared objects stronger than atomic registers, such as test&set bits, semaphores or readmodifywrite registers, and update them for the infinite process case.
A Bounded FirstIn, FirstEnabled Solution to the lExclusion Problem
 ACM Transactions on Programming Languages and Systems
, 1990
"... This paper presents a solution to the firstcome, firstenabled `exclusion problem of [?]. Unlike the solution in [?], this solution does not use powerful readmodifywrite synchronization primitives, and requires only bounded shared memory. Use of the concurrent timestamp system of [?] is key in s ..."
Abstract

Cited by 4 (0 self)
 Add to MetaCart
This paper presents a solution to the firstcome, firstenabled `exclusion problem of [?]. Unlike the solution in [?], this solution does not use powerful readmodifywrite synchronization primitives, and requires only bounded shared memory. Use of the concurrent timestamp system of [?] is key in solving the problem within bounded shared memory. Categories and Subject Descriptors: D.4.1 [Operating Systems]: Process ManagementMutual
Group Mutual Exclusion in Token Rings
 In SIROCCO 2001, The 8th International Colloquium On Structural Information and Communication Complexity Proceedings
, 2001
"... The group mutual exclusion (GME) problem was introduced by Joung [6]. The GME solution allows n processes to share m mutually exclusive resources. We first present a group mutual exclusion algorithm (Algorithm GME ) for anonymous token rings. The space requirement and the size of messages of thi ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
The group mutual exclusion (GME) problem was introduced by Joung [6]. The GME solution allows n processes to share m mutually exclusive resources. We first present a group mutual exclusion algorithm (Algorithm GME ) for anonymous token rings. The space requirement and the size of messages of this algorithm depend only on the number of shared resources (O(logm) bits). So, the proposed algorithm solves the problem suggested in [7], which is to obtain a solution using messages of bounded size. All costs related to the time depend on n. We then present two variations of Algorithm GM E . We design the second algorithm (Algorithm mGME) such that its cost depends mainly on the m instead of n. The third algorithm (Algorithm nmGME ) is a general algorithm which takes advantage of the lowest value between n and m. Keywords Distributed algorithms, group mutual exclusion, mutual exclusion. 1
A Note on Weighted Distributed MatchMaking
 Mathematical Systems Theory
, 1992
"... This is a preliminary draft version of [Mathematical Systems Theory, 25(1992), 123140]. In many distributed computing environments, processes are concurrently executed by nodes in a storeandforward network. Distributed control issues as diverse as nameserver, mutual exclusion and replicated data ..."
Abstract

Cited by 3 (2 self)
 Add to MetaCart
This is a preliminary draft version of [Mathematical Systems Theory, 25(1992), 123140]. In many distributed computing environments, processes are concurrently executed by nodes in a storeandforward network. Distributed control issues as diverse as nameserver, mutual exclusion and replicated data management, involve making matches between processes. The generic paradigm is a formal problem called "distributed matchmaking". We define multidimensional and weighted versions, and the relations between the two, and develop a very general method to prove lower bounds on the complexity as a tradeoff between number of messages and "distributedness ". The resulting lower bounds are tight in all cases we have examined. We present a successstop version of distributed matchmaking that is analysed in terms of a weight distribution that in all cases results in approximately halving the (expected) number of messages required in the corresponding strategy that does not use these weights. 1. I...
SelfStabilizing Timestamps
 Theoretical Computer Science
, 2001
"... The problem of implementing selfstabilizing timestamps with bounded values is investigated and a solution is found which is applied to the ` exclusion problem and to the Multiwriter Atomic Register problem. Thus we get selfstabilizing solutions to these two wellknown problems. A new type of ..."
Abstract

Cited by 2 (1 self)
 Add to MetaCart
The problem of implementing selfstabilizing timestamps with bounded values is investigated and a solution is found which is applied to the ` exclusion problem and to the Multiwriter Atomic Register problem. Thus we get selfstabilizing solutions to these two wellknown problems. A new type of weak timestamps is identied here, and some evidence is brought to show its usefulness. 1 Preface Messages are often timestamped. In a fax, the timestamp includes the date and exact time of the day, and in a book only the publication year, but in all cases this information guides the reader in choosing and processing the data. The antiquarian may choose the oldest book, and the student the newest edition, but in the general timestamp protocol the reader (called scanner) returns all the messages in their issuing order. 1 Timestamps may appear in conjunction with messages (\timestamped messages "), but timestamps may also appear alone in pure form, for example as numbers distributed to customers waiting for a certain service. We shall deal here with timestamped messages which are clearly more general, since by setting their data eld to the null value pure timestamps can be derived. Two well known problems will accompany our discussion: the `exclusion problem to illustrate pure timestamps, and the multiplewriter atomic register problem to illustrate timestamped messages. Another distinction is between unbounded timestamps (such as the natural numbers) and bounded timestamps which should achieve the same eect but with a bounded set of values. The classical Bakery Algorithm of Lamport (a 1 That the use of timestamps is old and natural is illustrated by ancient ostracons (about 800 BC) from Sumeria which show how oerings of vine and oil were marked by date, place of ori...
Some Modifications of the Tournament Algorithm for the Mutual Exclusion Problem
, 1999
"... Introduction Mutual exclusi) i a problem ofmanagiO access to asiz(4 iz(4]fiz)4z resource that can only support one user at ati/I An earlyalgori/) for the mutual exclusic problem was proposed byDiL4z(] [6]. TheorizOP/ versi/ of the DiPL//]fiz algori/] was presented on a modelassumiI a shared memory ..."
Abstract
 Add to MetaCart
Introduction Mutual exclusi) i a problem ofmanagiO access to asiz(4 iz(4]fiz)4z resource that can only support one user at ati/I An earlyalgori/) for the mutual exclusic problem was proposed byDiL4z(] [6]. TheorizOP/ versi/ of the DiPL//]fiz algori/] was presented on a modelassumiI a shared memorywio atomi read and wri] operatiP]fi The DiF//z/]fi algoriz/ guarantees mutualexclusi/] buti does not guaranteehi]z/PO el faiz(O44 SubsequentalgorifizF areie]z vements on the Di4FLz]fiz algori]fi by guaranteein fain]F to the die]z( t users [15], [16], weakeniF the type of shared memory [1][3], [5], [7][10], ordescrifiFL algorifiFL i more clearly speciFO shared memory models [9], [12], [13]. The tournament algori]fi for the mutual exclusic problemi orim]FPI( from the tournament protocol of Peterson andFi]FI [16]. Si]. the shared memory model usedi the tournament protocoli not clear enough to analyzeil runni] tini we use the tournament algorifi( descri ed i terms of I/O automatai t
Space and Time Efficient SelfStabilizing lExclusion in Tree Networks
, 2000
"... We propose a selfstabilizing `exclusion algorithm in rooted tree networks. The ` exclusion problem is a generalization of the mutual exclusion problemwe allow ` (` 1) processors, instead of 1, to use a shared resource. The algorithm is semiuniform and its space requirement is (` + 3) r states ..."
Abstract
 Add to MetaCart
We propose a selfstabilizing `exclusion algorithm in rooted tree networks. The ` exclusion problem is a generalization of the mutual exclusion problemwe allow ` (` 1) processors, instead of 1, to use a shared resource. The algorithm is semiuniform and its space requirement is (` + 3) r states (or dlog((` + 3) r )e bits) for the root r, 3 p states (or dlog(3 p )e bits) for an internal processor p, and 2 states (or 1 bit) for a leaf processor, where p is the degree of processor p. Our algorithm is unique in the sense that this is the rst `exclusion algorithm on trees, whose space requirement is independant of the size of the network for any processor, and is independent of ` for all processors except the root. The stabilization time of the algorithm is only O(` + h) rounds, where h is the height of the tree.