Results 1 -
5 of
5
Crash fault detection in celerating environments
- In Intl. Par. and Distrib. Proc. Symp
, 2009
"... Failure detectors are a service that provides (ap-proximate) information about process crashes in a dis-tributed system. The well-known “eventually perfect” failure detector, 3P, has been implemented in partially synchronous systems with unknown upper bounds on message delay and relative process spe ..."
Abstract
-
Cited by 4 (1 self)
- Add to MetaCart
(Show Context)
Failure detectors are a service that provides (ap-proximate) information about process crashes in a dis-tributed system. The well-known “eventually perfect” failure detector, 3P, has been implemented in partially synchronous systems with unknown upper bounds on message delay and relative process speeds. However, previous implementations have overlooked an important subtlety with respect to measuring the passage of time in “celerating ” environments, in which absolute process speeds can continually increase or decrease while main-taining bounds on relative process speeds. Existing im-plementations either use action clocks, which fail in ac-celerating environments, or use real-time clocks, which fail in decelerating environments. We propose the use of bichronal clocks, which are a composition of action clocks and real-time clocks. Our solution can be read-ily adopted to make existing implementations of 3P ro-bust to process celeration, which can result from hard-ware upgrades, server overloads, denial-of-service at-tacks, and other system volatilities. 1
Efficient and Robust Local Mutual Exclusion in Mobile Ad Hoc Networks (Extended Abstract)
"... This paper presents two algorithms for the local mutual exclusion problem, an extension of the dining philosophers problem for mobile ad hoc networks. A solution to this problem allows nodes that are currently geographically close to obtain exclusive access to a resource. The algorithms exhibit diff ..."
Abstract
-
Cited by 2 (0 self)
- Add to MetaCart
This paper presents two algorithms for the local mutual exclusion problem, an extension of the dining philosophers problem for mobile ad hoc networks. A solution to this problem allows nodes that are currently geographically close to obtain exclusive access to a resource. The algorithms exhibit different tradeoffs between response time and failure locality (the size of the neighborhood adversely affected by a node crash). The first algorithm has two variations, one of which has response time that depends very weakly on the number of nodes in the entire system and is polynomial in the maximum number of neighboring nodes; the failure locality, although not optimal, is small and grows very slowly with system size. The second algorithm has optimal failure locality and response time that is quadratic in the number of nodes. A pleasing aspect of this algorithm is that, when run in a system with no node movement, it has linear response time, improving on previous results for static algorithms with optimal failure locality. 1
Wait-Free Stabilizing Dining . . .
"... Dining philosophers is a scheduling paradigm that determines when processes in a distributed system should execute certain sections of their code so that processes do not execute ‘conflicting’ code sections concurrently, for some application-dependent notion of a ‘conflict’. Designing a stabilizing ..."
Abstract
- Add to MetaCart
Dining philosophers is a scheduling paradigm that determines when processes in a distributed system should execute certain sections of their code so that processes do not execute ‘conflicting’ code sections concurrently, for some application-dependent notion of a ‘conflict’. Designing a stabilizing dining algorithm for shared-memory systems subject to process crashes presents an interesting challenge: classic stabilization relies on all processes continuing to execute actions forever, an assumption which is violated when crash failures are considered. We present a dining algorithm that is both wait-free (tolerates any number of crashes) and is pseudo-stabilizing. Our algorithm works in an asynchronous system in which processes communicate via shared regular registers and have access to the eventually perfect failure detector ♦P. Furthermore, with a stronger failure detector, the solution becomes wait-free and self-stabilizing. To our knowledge, this is the first such algorithm. Prior results show that ♦P is necessary for wait-freedom.
Eventually Perfect Failure Detectors using ADD Channels
"... Abstract. We present a novel implementation of the eventually perfect failure detector (✸P) from the original hierarchy of Chandra-Toueg oracles. Previous implementations of ✸P have assumed models of partial synchrony where point-to-point message delay is bounded and/or communication is reliable. We ..."
Abstract
- Add to MetaCart
(Show Context)
Abstract. We present a novel implementation of the eventually perfect failure detector (✸P) from the original hierarchy of Chandra-Toueg oracles. Previous implementations of ✸P have assumed models of partial synchrony where point-to-point message delay is bounded and/or communication is reliable. We show how to implement this important oracle under even weaker assumptions using Average Delayed/Dropped (ADD) channels. Briefly, all messages sent on an ADD channel are privileged or non-privileged. All non-privileged messages can be arbitrarily delayed or even dropped. For each run, however, there exists an unknown window size w, and two unknown upper-bounds d and r, where d bounds the average delay of the last w privileged messages, and r bounds the ratio of non-privileged messages to privileged messages per window.
Stabilizing Dining with Failure Locality 1
"... Abstract. The dining philosophers problem, or simply dining, is a fun-damental distributed resource allocation problem. We propose two al-gorithms for solving stabilizing dining with failure locality 1 in asyn-chronous shared-memory systems with regular registers. Since this prob-lem cannot be solve ..."
Abstract
- Add to MetaCart
(Show Context)
Abstract. The dining philosophers problem, or simply dining, is a fun-damental distributed resource allocation problem. We propose two al-gorithms for solving stabilizing dining with failure locality 1 in asyn-chronous shared-memory systems with regular registers. Since this prob-lem cannot be solved in pure asynchrony, we augment the shared-memory system with failure detectors. Specifically, we introduce the local anony-mous eventually perfect failure detector?3P1, and show that this failure detector is sufficient to solve the problem at hand. 1
