Several recent cloud-backed storage systems advocates the composition of a number of cloud services for improving performance and fault tolerance (e.g., [1, 3, 4]). An interesting aspect of these compositions is that the consistency guarantees they provide depend on the consistency of such base services, which are normally different.
{"title":"On the consistency of heterogeneous composite objects","authors":"A. Bessani, Ricardo Mendes, Tiago Oliveira","doi":"10.1145/2745947.2746687","DOIUrl":"https://doi.org/10.1145/2745947.2746687","url":null,"abstract":"Several recent cloud-backed storage systems advocates the composition of a number of cloud services for improving performance and fault tolerance (e.g., [1, 3, 4]). An interesting aspect of these compositions is that the consistency guarantees they provide depend on the consistency of such base services, which are normally different.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121464565","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}
Modern internet applications require scalability to millions of clients, response times in the tens of milliseconds, and availability in the presence of partitions, hardware faults and even disasters. To obtain these requirements, applications are usually geo-replicated across several data centres (DCs) spread throughout the world, providing clients with fast access to nearby DCs and fault-tolerance in case of a DC out-age. Using multiple replicas also has disadvantages, not only does this incur extra storage, bandwidth and hardware costs, but programming these systems becomes more difficult. To address the additional hardware costs, data is often partially replicated, meaning that only certain DCs will keep a copy of certain data, for example in a key-value store it may only store values corresponding to a portion of the keys. Additionally, to address the issue of programming these systems, consistency protocols are run on top ensuring different guarantees for the data, but as shown by the CAP theorem, strong consistency, availability, and partition tolerance cannot be ensured at the same time. For many applications availability is paramout, thus strong consistency is exchanged for weaker consistencies allowing concurrent writes like causal consistency. Unfortunately these protocols are not designed with partial replication in mind and either end up not supporting it or do so in an inefficient manner. In this work we will look at why this happens and propose a protocol designed to support partial replication under causal consistency more efficiently.
{"title":"Designing a causally consistent protocol for geo-distributed partial replication","authors":"Tyler Crain, M. Shapiro","doi":"10.1145/2745947.2745953","DOIUrl":"https://doi.org/10.1145/2745947.2745953","url":null,"abstract":"Modern internet applications require scalability to millions of clients, response times in the tens of milliseconds, and availability in the presence of partitions, hardware faults and even disasters. To obtain these requirements, applications are usually geo-replicated across several data centres (DCs) spread throughout the world, providing clients with fast access to nearby DCs and fault-tolerance in case of a DC out-age. Using multiple replicas also has disadvantages, not only does this incur extra storage, bandwidth and hardware costs, but programming these systems becomes more difficult. To address the additional hardware costs, data is often partially replicated, meaning that only certain DCs will keep a copy of certain data, for example in a key-value store it may only store values corresponding to a portion of the keys. Additionally, to address the issue of programming these systems, consistency protocols are run on top ensuring different guarantees for the data, but as shown by the CAP theorem, strong consistency, availability, and partition tolerance cannot be ensured at the same time. For many applications availability is paramout, thus strong consistency is exchanged for weaker consistencies allowing concurrent writes like causal consistency. Unfortunately these protocols are not designed with partial replication in mind and either end up not supporting it or do so in an inefficient manner. In this work we will look at why this happens and propose a protocol designed to support partial replication under causal consistency more efficiently.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128214216","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 this paper, we introduce a technique that can be used by distributed transactional protocols to reduce the vulnerability window of transactions. For this purpose, we propose a so far unexplored (to the best of our knowledge) usage of hybrid clocks. On one hand, loosely synchronized physical clocks are used to maximize the freshness of the snapshots used by transactions to read. On the other hand, logical clocks are used to reduce the extent to which the snapshot of update transactions is advanced upon their commit. We claim that the joint usage of these two techniques can potentially reduce the abort rate in comparison to previous protocols such as Clock-SI, GMU, and SCORe.
{"title":"Reducing the vulnerability window in distributed transactional protocols","authors":"Manuel Bravo, P. Romano, L. Rodrigues, P. V. Roy","doi":"10.1145/2745947.2746688","DOIUrl":"https://doi.org/10.1145/2745947.2746688","url":null,"abstract":"In this paper, we introduce a technique that can be used by distributed transactional protocols to reduce the vulnerability window of transactions. For this purpose, we propose a so far unexplored (to the best of our knowledge) usage of hybrid clocks. On one hand, loosely synchronized physical clocks are used to maximize the freshness of the snapshots used by transactions to read. On the other hand, logical clocks are used to reduce the extent to which the snapshot of update transactions is advanced upon their commit. We claim that the joint usage of these two techniques can potentially reduce the abort rate in comparison to previous protocols such as Clock-SI, GMU, and SCORe.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132208859","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}
B. Holt, Irene Zhang, Dan R. K. Ports, M. Oskin, L. Ceze
Out of the many NoSQL databases in use today, some that provide simple data structures for records, such as Redis and MongoDB, are now becoming popular. Building applications out of these complex data types provides a way to communicate intent to the database system without sacrificing flexibility or committing to a fixed schema. Currently this capability is leveraged in limited ways, such as to ensure related values are co-located, or for atomic updates. There are many ways data types can be used to make databases more efficient that are not yet being exploited. We explore several ways of leveraging abstract data type (ADT) semantics in databases, focusing primarily on commutativity. Using a Twitter clone as a case study, we show that using commutativity can reduce transaction abort rates for high-contention, update-heavy workloads that arise in real social networks. We conclude that ADTs are a good abstraction for database records, providing a safe and expressive programming model with ample opportunities for optimization, making databases more safe and scalable.
{"title":"Claret: using data types for highly concurrent distributed transactions","authors":"B. Holt, Irene Zhang, Dan R. K. Ports, M. Oskin, L. Ceze","doi":"10.1145/2745947.2745951","DOIUrl":"https://doi.org/10.1145/2745947.2745951","url":null,"abstract":"Out of the many NoSQL databases in use today, some that provide simple data structures for records, such as Redis and MongoDB, are now becoming popular. Building applications out of these complex data types provides a way to communicate intent to the database system without sacrificing flexibility or committing to a fixed schema. Currently this capability is leveraged in limited ways, such as to ensure related values are co-located, or for atomic updates. There are many ways data types can be used to make databases more efficient that are not yet being exploited. We explore several ways of leveraging abstract data type (ADT) semantics in databases, focusing primarily on commutativity. Using a Twitter clone as a case study, we show that using commutativity can reduce transaction abort rates for high-contention, update-heavy workloads that arise in real social networks. We conclude that ADTs are a good abstraction for database records, providing a safe and expressive programming model with ample opportunities for optimization, making databases more safe and scalable.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116288614","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}
Causal consistency is the strongest consistency model under which low-latency and high-availability can be achieved. In the past few years, many causally consistent storage systems have been developed. The long-term goal of this initial work is to perform a deep study and comparison of the different implementations of causal consistency. We identify that protocols that provide causal consistency share the well-known DUR (deferred update replication) algorithmic structure and observe that existing implementations of causal consistency fall into a sub-category of DUR that we name A-DUR (Asynchronous-DUR). In this work, we present the A-DUR algorithmic structure, the pseudocode for the instantiation of two causally consistent protocols under the G-DUR framework, and describe the empirical study we intend to perform on causal consistency.
{"title":"An empirical perspective on causal consistency","authors":"Alejandro Z. Tomsic, Tyler Crain, M. Shapiro","doi":"10.1145/2745947.2745949","DOIUrl":"https://doi.org/10.1145/2745947.2745949","url":null,"abstract":"Causal consistency is the strongest consistency model under which low-latency and high-availability can be achieved. In the past few years, many causally consistent storage systems have been developed. The long-term goal of this initial work is to perform a deep study and comparison of the different implementations of causal consistency. We identify that protocols that provide causal consistency share the well-known DUR (deferred update replication) algorithmic structure and observe that existing implementations of causal consistency fall into a sub-category of DUR that we name A-DUR (Asynchronous-DUR). In this work, we present the A-DUR algorithmic structure, the pseudocode for the instantiation of two causally consistent protocols under the G-DUR framework, and describe the empirical study we intend to perform on causal consistency.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133639996","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 amount of data being processed in Data Centres (DCs) keeps growing at an enormous rate so that full replication may start being impractical. The application of replication between DCs is used to increase data availability in the presence of site failures and to reduce latency by accessing the data closely located, if possible. This means that replicating the data only in some of the DCs is becoming more critical in order to reduce the costs of keeping the data (weakly) consistent while maintaining high availability (scalability) and low access costs. When data read and write request patterns change, then deciding which data should be replicated and where needs to be made dynamically. Given that the problem of finding an optimal replication schema in a general network has been shown to be NP-complete for the static case, so it is unlikely that there exists a general algorithm for an optimal solution to the dynamic problem. We present here a new adaptive bio--inspired replication strategy, which is completely decentralised, adaptive, and event-driven, inspired on the Ant Colony algorithm.
{"title":"Adaptive strength geo-replication strategy","authors":"Amadeo Ascó Signes, Annette Bieniusa","doi":"10.1145/2745947.2745950","DOIUrl":"https://doi.org/10.1145/2745947.2745950","url":null,"abstract":"The amount of data being processed in Data Centres (DCs) keeps growing at an enormous rate so that full replication may start being impractical. The application of replication between DCs is used to increase data availability in the presence of site failures and to reduce latency by accessing the data closely located, if possible. This means that replicating the data only in some of the DCs is becoming more critical in order to reduce the costs of keeping the data (weakly) consistent while maintaining high availability (scalability) and low access costs. When data read and write request patterns change, then deciding which data should be replicated and where needs to be made dynamically. Given that the problem of finding an optimal replication schema in a general network has been shown to be NP-complete for the static case, so it is unlikely that there exists a general algorithm for an optimal solution to the dynamic problem. We present here a new adaptive bio--inspired replication strategy, which is completely decentralised, adaptive, and event-driven, inspired on the Ant Colony algorithm.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132451175","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 use cases for Conflict-free Replicated Data Types (CRDTs) that are studied in the literature are limited to collaborative editing applications and data stores. The communication protocols used to distribute replica updates in these scenarios are usually assumed to be some form of highly scalable gossip protocol. In this paper, a new type of application for CRDTs is introduced and studied: collaborative offline web applications. We demonstrate the feasibility of CRDTs in this scenario, and analyze the trade-offs of three existing communication protocols that can be employed for these applications.
{"title":"Collaborative offline web applications using conflict-free replicated data types","authors":"Santiago J. Castiñeira, Annette Bieniusa","doi":"10.1145/2745947.2745952","DOIUrl":"https://doi.org/10.1145/2745947.2745952","url":null,"abstract":"The use cases for Conflict-free Replicated Data Types (CRDTs) that are studied in the literature are limited to collaborative editing applications and data stores. The communication protocols used to distribute replica updates in these scenarios are usually assumed to be some form of highly scalable gossip protocol. In this paper, a new type of application for CRDTs is introduced and studied: collaborative offline web applications. We demonstrate the feasibility of CRDTs in this scenario, and analyze the trade-offs of three existing communication protocols that can be employed for these applications.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115145598","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}
A CRDT is a data type specially designed to allow multiple instances to be replicated and modified without coordination, while providing an automatic mechanism to merge concurrent updates that guarantee eventual consistency. In this paper we present a brief study of computational CRDTs, a class of CRDTs whose state is the result of a computation over the executed updates. We propose three generic designs that reduce the amount of information that each replica maintains and propagates for synchronizations. For each of the designs, we discuss the properties that the function being computed needs to satisfy.
{"title":"A study of CRDTs that do computations","authors":"David Navalho, S. Duarte, Nuno M. Preguiça","doi":"10.1145/2745947.2745948","DOIUrl":"https://doi.org/10.1145/2745947.2745948","url":null,"abstract":"A CRDT is a data type specially designed to allow multiple instances to be replicated and modified without coordination, while providing an automatic mechanism to merge concurrent updates that guarantee eventual consistency. In this paper we present a brief study of computational CRDTs, a class of CRDTs whose state is the result of a computation over the executed updates. We propose three generic designs that reduce the amount of information that each replica maintains and propagates for synchronizations. For each of the designs, we discuss the properties that the function being computed needs to satisfy.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128271383","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}
Cheng Li, J. Leitao, Allen Clement, Nuno M. Preguiça, R. Rodrigues
Replication has been widely adopted to build highly scalable services, but this goal is often compromised by the coordination required to ensure application-specific properties such as state convergence and invariant preservation. In this paper, we propose a principled mechanism to minimize coordination in replicated systems via the following components: a) a notion of restriction over pairs of operations, which captures the fact that the two operations must be ordered w.r.t. each other in any partial order; b) a generic consistency model which, given a set of restrictions, requires those restrictions to be met in all admissible partial orders; c) principles for identifying a minimal set of restrictions to ensure the above properties; and d) a coordination service that dynamically maps restrictions to the most efficient coordination protocols. Our preliminary experience with example applications shows that we are able to determine a minimal coordination strategy.
{"title":"Minimizing coordination in replicated systems","authors":"Cheng Li, J. Leitao, Allen Clement, Nuno M. Preguiça, R. Rodrigues","doi":"10.1145/2745947.2745955","DOIUrl":"https://doi.org/10.1145/2745947.2745955","url":null,"abstract":"Replication has been widely adopted to build highly scalable services, but this goal is often compromised by the coordination required to ensure application-specific properties such as state convergence and invariant preservation. In this paper, we propose a principled mechanism to minimize coordination in replicated systems via the following components: a) a notion of restriction over pairs of operations, which captures the fact that the two operations must be ordered w.r.t. each other in any partial order; b) a generic consistency model which, given a set of restrictions, requires those restrictions to be met in all admissible partial orders; c) principles for identifying a minimal set of restrictions to ensure the above properties; and d) a coordination service that dynamically maps restrictions to the most efficient coordination protocols. Our preliminary experience with example applications shows that we are able to determine a minimal coordination strategy.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133381252","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 propose Lasp, a novel programming model aimed to simplify correct, large-scale, distributed programming. Lasp leverages ideas from distributed dataflow programming extended with convergent data types. This provides support for computations where not all participants are online together at a given moment through Lasp's "convergent by design" applications. Lasp provides a familiar functional programming semantics, built on top of distributed systems infrastructure, targeted at the Erlang runtime system. The initial Lasp design presented in this report supports synchronization free programming using convergent data types. It combines the expressiveness of these data types together with powerful primitives for composing them. This design lets us write long-lived fault-tolerant distributed applications with non-monotonic behavior. We show how to implement one nontrivial large-scale application, the ad counter scenario from the SyncFree project.
{"title":"Lasp: a language for distributed, eventually consistent computations with CRDTs","authors":"Christopher S. Meiklejohn, P. V. Roy","doi":"10.1145/2745947.2745954","DOIUrl":"https://doi.org/10.1145/2745947.2745954","url":null,"abstract":"We propose Lasp, a novel programming model aimed to simplify correct, large-scale, distributed programming. Lasp leverages ideas from distributed dataflow programming extended with convergent data types. This provides support for computations where not all participants are online together at a given moment through Lasp's \"convergent by design\" applications. Lasp provides a familiar functional programming semantics, built on top of distributed systems infrastructure, targeted at the Erlang runtime system. The initial Lasp design presented in this report supports synchronization free programming using convergent data types. It combines the expressiveness of these data types together with powerful primitives for composing them. This design lets us write long-lived fault-tolerant distributed applications with non-monotonic behavior. We show how to implement one nontrivial large-scale application, the ad counter scenario from the SyncFree project.","PeriodicalId":332245,"journal":{"name":"Proceedings of the First Workshop on Principles and Practice of Consistency for Distributed Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2015-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134109537","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：