M. Aguilera, Tudor David, R. Guerraoui, Junxiong Wang
We present multiversion timestamp locking (MVTL), a new genre of multiversion concurrency control algorithms for serializable transactions. The key idea behind MVTL is simple: lock individual timestamps instead of locking objects. After presenting a generic MVTL algorithm, we demonstrate MVTL's expressiveness: we give several simple MVTL algorithms that address limitations of current multiversion schemes, by committing transactions that previous schemes would abort, by avoiding the problems of serial aborts or ghost aborts, and by offering a way to prioritize transactions that should not be aborted. We give evidence that, in practice, MVTL-based algorithms can outperform alternative concurrency control schemes.
{"title":"Locking Timestamps versus Locking Objects","authors":"M. Aguilera, Tudor David, R. Guerraoui, Junxiong Wang","doi":"10.1145/3212734.3212742","DOIUrl":"https://doi.org/10.1145/3212734.3212742","url":null,"abstract":"We present multiversion timestamp locking (MVTL), a new genre of multiversion concurrency control algorithms for serializable transactions. The key idea behind MVTL is simple: lock individual timestamps instead of locking objects. After presenting a generic MVTL algorithm, we demonstrate MVTL's expressiveness: we give several simple MVTL algorithms that address limitations of current multiversion schemes, by committing transactions that previous schemes would abort, by avoiding the problems of serial aborts or ghost aborts, and by offering a way to prioritize transactions that should not be aborted. We give evidence that, in practice, MVTL-based algorithms can outperform alternative concurrency control schemes.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114935613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper gives a brief presentation of a comprehensive study on the necessary and sufficient state space conditions for the deterministic naming task in the population protocol model. This problem is studied under various combinations of model assumptions: weak or global fairness, arbitrary or uniform initialization of agents, existence or absence of a distinguishable agent (arbitrarily initialized or not), possibility of breaking symmetry in pair-wise interactions (symmetric or asymmetric transitions). For each possible combination of these assumptions, either an impossibility is proven or the necessary exact number of states (per mobile agent) is determined and an appropriate space-optimal naming protocol is given.
{"title":"Brief Announcement: Space-Optimal Naming in Population Protocols","authors":"J. Burman, J. Beauquier, D. Sohier","doi":"10.1145/3212734.3212791","DOIUrl":"https://doi.org/10.1145/3212734.3212791","url":null,"abstract":"This paper gives a brief presentation of a comprehensive study on the necessary and sufficient state space conditions for the deterministic naming task in the population protocol model. This problem is studied under various combinations of model assumptions: weak or global fairness, arbitrary or uniform initialization of agents, existence or absence of a distinguishable agent (arbitrarily initialized or not), possibility of breaking symmetry in pair-wise interactions (symmetric or asymmetric transitions). For each possible combination of these assumptions, either an impossibility is proven or the necessary exact number of states (per mobile agent) is determined and an appropriate space-optimal naming protocol is given.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122439938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Distributed shared memory systems maintain multiple replicas of the shared memory registers. Maintaining causal consistency in such systems has received significant attention in the past. However, much of the previous literature focuses on em full replication wherein each replica stores a copy of all the registers in the shared memory. In this paper, we investigate causal consistency in partially replicated systems, wherein each replica may store only a subset of the shared data. To achieve causal consistency, it is necessary to ensure that, before an update is performed at any given replica, all causally preceding updates must also be performed. Achieving this goal requires some mechanism to track causal dependencies. In the context of full replication, this goal is often achieved using vector timestamps, with the number of vector elements being equal to the number of replicas. Building on the past work, this paper makes two key contributions: For a family of algorithms for maintaining causal consistency, we present necessary conditions on the metadata (which we refer as a em timestamp ) that must be maintained by each replica. We present an algorithm for achieving causal consistency using a timestamp that matches one of the necessary conditions referred above, thus showing that the condition is necessary and sufficient both.
{"title":"Brief Announcement: Partially Replicated Causally Consistent Shared Memory","authors":"Zhuolun Xiang, N. Vaidya","doi":"10.1145/3212734.3212790","DOIUrl":"https://doi.org/10.1145/3212734.3212790","url":null,"abstract":"Distributed shared memory systems maintain multiple replicas of the shared memory registers. Maintaining causal consistency in such systems has received significant attention in the past. However, much of the previous literature focuses on em full replication wherein each replica stores a copy of all the registers in the shared memory. In this paper, we investigate causal consistency in partially replicated systems, wherein each replica may store only a subset of the shared data. To achieve causal consistency, it is necessary to ensure that, before an update is performed at any given replica, all causally preceding updates must also be performed. Achieving this goal requires some mechanism to track causal dependencies. In the context of full replication, this goal is often achieved using vector timestamps, with the number of vector elements being equal to the number of replicas. Building on the past work, this paper makes two key contributions: For a family of algorithms for maintaining causal consistency, we present necessary conditions on the metadata (which we refer as a em timestamp ) that must be maintained by each replica. We present an algorithm for achieving causal consistency using a timestamp that matches one of the necessary conditions referred above, thus showing that the condition is necessary and sufficient both.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128243200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Session details: Session 2B: Routing and Leader Election","authors":"M. Ghaffari","doi":"10.1145/3258700","DOIUrl":"https://doi.org/10.1145/3258700","url":null,"abstract":"","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129197461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Distributed systems underlying large-scale Internet services often guarantee immediate availability and tolerate network failures at the expense of providing only weak data consistency guarantees. This is compensated for by new programming constructs, such as replicated data types (aka CRDTs) and novel forms of transactions. Navigating the spectrum of possible consistency models and programming constructs is far from trivial. This tutorial surveys recent developments that help in this: formal definitions of consistency model semantics and methods for reasoning about how the weakness of consistency models affects the correctness of applications using them.
{"title":"Tutorial: Consistency Choices in Modern Distributed Systems","authors":"Alexey Gotsman","doi":"10.1145/3212734.3212800","DOIUrl":"https://doi.org/10.1145/3212734.3212800","url":null,"abstract":"Distributed systems underlying large-scale Internet services often guarantee immediate availability and tolerate network failures at the expense of providing only weak data consistency guarantees. This is compensated for by new programming constructs, such as replicated data types (aka CRDTs) and novel forms of transactions. Navigating the spectrum of possible consistency models and programming constructs is far from trivial. This tutorial surveys recent developments that help in this: formal definitions of consistency model semantics and methods for reasoning about how the weakness of consistency models affects the correctness of applications using them.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113979419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
H. Attiya, Armando Castañeda, Danny Hendler, Matthieu Perrin
A long-standing open question has been whether lock-freedom and wait-freedom are fundamentally different progress conditions, namely, can the former be provided in situations where the latter cannot? This paper answers the question in the affirmative, by proving that there are objects with lock-free implementations, but without wait-free implementations-using objects of any finite power. We precisely define an object called n-process long-lived approximate agreement (n-LLAA), in which two sets of processes associated with two sides, 0 or 1, need to decide on a sequence of increasingly closer outputs. We prove that 2-LLAA has a lock-free implementation using reads and writes only, while n-LLAA has a lock-free implementation using reads, writes and (n - 1)-process consensus objects. In contrast, we prove that there is no wait-free implementation of the n-LLAA object using reads, writes and specific (n - 1)-process consensus objects, called (n - 1)-window registers.
{"title":"Separating Lock-Freedom from Wait-Freedom","authors":"H. Attiya, Armando Castañeda, Danny Hendler, Matthieu Perrin","doi":"10.1145/3212734.3212739","DOIUrl":"https://doi.org/10.1145/3212734.3212739","url":null,"abstract":"A long-standing open question has been whether lock-freedom and wait-freedom are fundamentally different progress conditions, namely, can the former be provided in situations where the latter cannot? This paper answers the question in the affirmative, by proving that there are objects with lock-free implementations, but without wait-free implementations-using objects of any finite power. We precisely define an object called n-process long-lived approximate agreement (n-LLAA), in which two sets of processes associated with two sides, 0 or 1, need to decide on a sequence of increasingly closer outputs. We prove that 2-LLAA has a lock-free implementation using reads and writes only, while n-LLAA has a lock-free implementation using reads, writes and (n - 1)-process consensus objects. In contrast, we prove that there is no wait-free implementation of the n-LLAA object using reads, writes and specific (n - 1)-process consensus objects, called (n - 1)-window registers.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134228379","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
When systems fail, logs are frequently the only evidence available for underlying fault diagnosis. Consequently, the quality of logs-how well system faults can be reflected by these log messages, is of significant importance. To improve the quality of logs, we propose a novel log enhancement approach which automatically identifies logging points that reflect anomalous behavior during system fault time. We further evaluate our approach with three popular open source projects. Results show that it can significantly improve over 50% accuracy of automatic fault diagnosis on average.
{"title":"Brief Announcement: Automatic Log Enhancement for Fault Diagnosis","authors":"Tong Jia, Ying Li, Zhonghai Wu","doi":"10.1145/3212734.3212784","DOIUrl":"https://doi.org/10.1145/3212734.3212784","url":null,"abstract":"When systems fail, logs are frequently the only evidence available for underlying fault diagnosis. Consequently, the quality of logs-how well system faults can be reflected by these log messages, is of significant importance. To improve the quality of logs, we propose a novel log enhancement approach which automatically identifies logging points that reflect anomalous behavior during system fault time. We further evaluate our approach with three popular open source projects. Results show that it can significantly improve over 50% accuracy of automatic fault diagnosis on average.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129451826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The physical (or SINR) model of wireless communication is more intricate than radio networks and still not well understood. If two neighbors of a node are transmitting, the node may be able to decode one of the transmissions, depending on the relative nearness of the transmitters. Thus, even the lack of proper reception carries indirect information. We explore here the power of such indirect signaling to infer the approximate topology of the network. In particular, we obtain a polylogarithmic time algorithm to compute a backbone: a set of nodes of constant density that dominates every ε-neighborhood. A backbone has wide utility for information dissemination, functioning as a sparse spanner. It also leads to fast broadcast, running in O(Diameter) time after a polylogarithmic precomputation, which previously was only known when additional features such as carrier sense, collision detection, geometric coordinates, or power control were available.
{"title":"Leveraging Indirect Signaling for Topology Inference and Fast Broadcast","authors":"M. Halldórsson, Tigran Tonoyan","doi":"10.1145/3212734.3212766","DOIUrl":"https://doi.org/10.1145/3212734.3212766","url":null,"abstract":"The physical (or SINR) model of wireless communication is more intricate than radio networks and still not well understood. If two neighbors of a node are transmitting, the node may be able to decode one of the transmissions, depending on the relative nearness of the transmitters. Thus, even the lack of proper reception carries indirect information. We explore here the power of such indirect signaling to infer the approximate topology of the network. In particular, we obtain a polylogarithmic time algorithm to compute a backbone: a set of nodes of constant density that dominates every ε-neighborhood. A backbone has wide utility for information dissemination, functioning as a sparse spanner. It also leads to fast broadcast, running in O(Diameter) time after a polylogarithmic precomputation, which previously was only known when additional features such as carrier sense, collision detection, geometric coordinates, or power control were available.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129676461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We briefly describe an efficient lock-free concurrent stack design with tunable and tenable relaxed semantics to allow for better performance. The design is tunable and allow for a continuous monotonic trade of weaker semantics for better throughput performance. Concurrent stacks have an inherent scalability bottleneck due to their single access point for both their operations. Elimination and semantics relaxation have been proposed in the literature to address this problem. Semantics relaxation has the potential to reach monotonically very high throughput by continuously trading relaxation for throughput. Previous solutions could not fully leverage this potential. We suggest a new two dimensional design that can achieve this by exploiting disjoint access parallelism in one dimension and locality in the other within tight accuracy bounds. The behaviour of the algorithm is tightly bound. We compare experimentally to previous work, with respect to throughput and relaxed behaviour observed, on different relaxation and concurrency settings. The experimental evaluation shows that our algorithm significantly outperform all other algorithms in terms of performance, also maintain better accuracy in contrast to other designs with relaxed semantics.
{"title":"Brief Announcement: 2D-Stack -- A Scalable Lock-Free Stack Design that Continuously Relaxes Semantics for Better Performance","authors":"A. Rukundo, A. Atalar, P. Tsigas","doi":"10.1145/3212734.3212794","DOIUrl":"https://doi.org/10.1145/3212734.3212794","url":null,"abstract":"We briefly describe an efficient lock-free concurrent stack design with tunable and tenable relaxed semantics to allow for better performance. The design is tunable and allow for a continuous monotonic trade of weaker semantics for better throughput performance. Concurrent stacks have an inherent scalability bottleneck due to their single access point for both their operations. Elimination and semantics relaxation have been proposed in the literature to address this problem. Semantics relaxation has the potential to reach monotonically very high throughput by continuously trading relaxation for throughput. Previous solutions could not fully leverage this potential. We suggest a new two dimensional design that can achieve this by exploiting disjoint access parallelism in one dimension and locality in the other within tight accuracy bounds. The behaviour of the algorithm is tightly bound. We compare experimentally to previous work, with respect to throughput and relaxed behaviour observed, on different relaxation and concurrency settings. The experimental evaluation shows that our algorithm significantly outperform all other algorithms in terms of performance, also maintain better accuracy in contrast to other designs with relaxed semantics.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117146044","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the uniformity testing problem, we are given access to samples from some unknown distribution μ on a fixed domain set1,..,n , and our goal is to distinguish the case where μ is the uniform distribution from the case where μ is ε-far from uniform in L_1 distance. Centralized uniformity testing has been extensively studied, and it is known that Θ(√ /ε^2) samples are necessary and sufficient. In this paper we study distributed uniformity testing : in a network of k nodes, each node i has access to s_i samples from the underlying distribution μ. Our goal is to test uniformity, while minimizing the number of samples per node, as well as the running time. We consider several distributed models: the ŁOCAL model, the CONGEST model, and a 0-round model where nodes cannot communicate with each other at all. We give upper bounds for each model, and a lower bound for the 0-round model. The key to our results is analyzing the centralized uniformity-testing problem in an unusual error regime, for which we give new upper and lower bounds.
{"title":"Distributed Uniformity Testing","authors":"O. Fischer, Uri Meir, R. Oshman","doi":"10.1145/3212734.3212772","DOIUrl":"https://doi.org/10.1145/3212734.3212772","url":null,"abstract":"In the uniformity testing problem, we are given access to samples from some unknown distribution μ on a fixed domain set1,..,n , and our goal is to distinguish the case where μ is the uniform distribution from the case where μ is ε-far from uniform in L_1 distance. Centralized uniformity testing has been extensively studied, and it is known that Θ(√ /ε^2) samples are necessary and sufficient. In this paper we study distributed uniformity testing : in a network of k nodes, each node i has access to s_i samples from the underlying distribution μ. Our goal is to test uniformity, while minimizing the number of samples per node, as well as the running time. We consider several distributed models: the ŁOCAL model, the CONGEST model, and a 0-round model where nodes cannot communicate with each other at all. We give upper bounds for each model, and a lower bound for the 0-round model. The key to our results is analyzing the centralized uniformity-testing problem in an unusual error regime, for which we give new upper and lower bounds.","PeriodicalId":198284,"journal":{"name":"Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123681578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：