Synchronisation is a bottleneck in shared memory concurrent programs. In this paper, we present Mergeable Data Types (MDTs) for relaxing synchronisation in shared memory system analogously to CRDTs in replicated systems. We present the key properties of MDTs and discuss the design of several MDTs.
{"title":"Highly-scalable concurrent objects","authors":"Deepthi Devaki Akkoorath, Annette Bieniusa","doi":"10.1145/2911151.2911158","DOIUrl":"https://doi.org/10.1145/2911151.2911158","url":null,"abstract":"Synchronisation is a bottleneck in shared memory concurrent programs. In this paper, we present Mergeable Data Types (MDTs) for relaxing synchronisation in shared memory system analogously to CRDTs in replicated systems. We present the key properties of MDTs and discuss the design of several MDTs.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128934229","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}
Valter Balegas, S. Duarte, Carla Ferreira, Nuno M. Preguiça, R. Rodrigues
This paper focuses on the problem of implementing web applications on top of weakly consistent geo-replicated systems. Several techniques, such as CRDTs, have been proposed to achieve state convergence on a per-object and per-data type basis. However, that does not guarantee application correctness, as convergence rules applied individually at each object may lead to an invalid state. We advocate that it is possible to address these problems and implement correct applications under weak consistency. To that end, it is necessary to combine CRDTs with novel semantics, judiciously select the CRDTs that are used by applications, and transform application operations to guarantee that convergence rules, applied on a per-object basis, always lead to valid application states. Achieving this is complex and requires tools to help programmers tame the complexity of programming on top of weak consistency and make the technology more accessible. In the presentation of this work we make a demonstration of a prototype tool that is capable of detecting concurrency conflicts on applications and propose transformations to make them conflict-free.
{"title":"Making weak consistency great again","authors":"Valter Balegas, S. Duarte, Carla Ferreira, Nuno M. Preguiça, R. Rodrigues","doi":"10.1145/2911151.2911167","DOIUrl":"https://doi.org/10.1145/2911151.2911167","url":null,"abstract":"This paper focuses on the problem of implementing web applications on top of weakly consistent geo-replicated systems. Several techniques, such as CRDTs, have been proposed to achieve state convergence on a per-object and per-data type basis. However, that does not guarantee application correctness, as convergence rules applied individually at each object may lead to an invalid state. We advocate that it is possible to address these problems and implement correct applications under weak consistency. To that end, it is necessary to combine CRDTs with novel semantics, judiciously select the CRDTs that are used by applications, and transform application operations to guarantee that convergence rules, applied on a per-object basis, always lead to valid application states. Achieving this is complex and requires tools to help programmers tame the complexity of programming on top of weak consistency and make the technology more accessible. In the presentation of this work we make a demonstration of a prototype tool that is capable of detecting concurrency conflicts on applications and propose transformations to make them conflict-free.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131495498","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}
Many machine learning (ML) models are of a stochastic nature. We aim to combine the principles of weak consistency with large scale distributed machine learning. We see interesting opportunities in this domain in (1) perceiving parallel ML algorithms based on model replication as a "collaborative task" where local progress on models is instantaneously exchanged and by (2) making this exchange more efficient by exploiting the underlying stochastic nature. Based on this motivation, we extend the notion of consistency for replicated objects with intrinsic stochastic structure and introduce harmonization as the reconciliation principle to enable efficient consistency maintenance of these objects. We present as a concrete application the harmonization of replicated ML models.
{"title":"Weak consistency and stochastic environments: harmonization of replicated machine learning models","authors":"T. Herb, T. Jungnickel, Christoph Alt","doi":"10.1145/2911151.2911161","DOIUrl":"https://doi.org/10.1145/2911151.2911161","url":null,"abstract":"Many machine learning (ML) models are of a stochastic nature. We aim to combine the principles of weak consistency with large scale distributed machine learning. We see interesting opportunities in this domain in (1) perceiving parallel ML algorithms based on model replication as a \"collaborative task\" where local progress on models is instantaneously exchanged and by (2) making this exchange more efficient by exploiting the underlying stochastic nature. Based on this motivation, we extend the notion of consistency for replicated objects with intrinsic stochastic structure and introduce harmonization as the reconciliation principle to enable efficient consistency maintenance of these objects. We present as a concrete application the harmonization of replicated ML models.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128489273","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 order to allow offline functionality of (mobile) web applications, state needs to be optimistically replicated and synchronized whenever connection is re-established. We present a programming language solution that provides replicated application state in a cloud-client setting. It exposes first-class replicated objects to the programmer, which are automatically replicated to all clients and locally accessible. By executing all method invocations in the same order on all sites, as determined by the cloud, consistency is attained. We call this serializable eventual consistency. The model encapsulates tentative operation buffering, reordering and propagation and provides primitives to reason about tentative state.
{"title":"Serializable eventual consistency: consistency through object method replay","authors":"Tim Coppieters, W. Meuter, S. Burckhardt","doi":"10.1145/2911151.2911164","DOIUrl":"https://doi.org/10.1145/2911151.2911164","url":null,"abstract":"In order to allow offline functionality of (mobile) web applications, state needs to be optimistically replicated and synchronized whenever connection is re-established. We present a programming language solution that provides replicated application state in a cloud-client setting. It exposes first-class replicated objects to the programmer, which are automatically replicated to all clients and locally accessible. By executing all method invocations in the same order on all sites, as determined by the cloud, consistency is attained. We call this serializable eventual consistency. The model encapsulates tentative operation buffering, reordering and propagation and provides primitives to reason about tentative state.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117073094","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}
Replication is a key technique for providing both fault tolerance and availability in distributed systems. However, managing replicated state, and ensuring that these replicas remain consistent, is a non trivial task, in particular in scenarios where replicas can reside on the client-side, as clients might have unreliable communication channels and hence, exhibit highly dynamic communication patterns. One way to simplify this task is to resort to CRDTs, which are data types that enable replication and operation over replicas with no coordination, ensuring eventual state convergence when these replicas are synchronized. However, when the communication patters, and therefore synchronization patterns, are highly dynamic, existing designs of CRDTs might incur in excessive communication overhead. To address those scenarios, in this paper we propose a new design for CRDTs which we call Δ-CRDT, and experimentally show that under dynamic communication patters, this novel design achieves better network utilization than existing alternatives.
{"title":"Δ-CRDTs: making δ-CRDTs delta-based","authors":"A. Linde, J. Leitao, Nuno M. Preguiça","doi":"10.1145/2911151.2911163","DOIUrl":"https://doi.org/10.1145/2911151.2911163","url":null,"abstract":"Replication is a key technique for providing both fault tolerance and availability in distributed systems. However, managing replicated state, and ensuring that these replicas remain consistent, is a non trivial task, in particular in scenarios where replicas can reside on the client-side, as clients might have unreliable communication channels and hence, exhibit highly dynamic communication patterns. One way to simplify this task is to resort to CRDTs, which are data types that enable replication and operation over replicas with no coordination, ensuring eventual state convergence when these replicas are synchronized. However, when the communication patters, and therefore synchronization patterns, are highly dynamic, existing designs of CRDTs might incur in excessive communication overhead. To address those scenarios, in this paper we propose a new design for CRDTs which we call Δ-CRDT, and experimentally show that under dynamic communication patters, this novel design achieves better network utilization than existing alternatives.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116877193","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}
Mahsa Najafzadeh, Alexey Gotsman, Hongseok Yang, Carla Ferreira, M. Shapiro
Designers of a replicated database face a vexing choice between strong consistency, which ensures certain application invariants but is slow and fragile, and asynchronous replication, which is highly available and responsive, but exposes the programmer to unfamiliar behaviours. To bypass this conundrum, recent research has studied hybrid consistency models, in which updates are asynchronous by default, but synchronisation is available upon request. To help programmers exploit hybrid consistency, we propose the first static analysis tool for proving integrity invariants of applications using databases with hybrid consistency models. This allows a programmer to find minimal consistency guarantees sufficient for application correctness.
{"title":"The CISE tool: proving weakly-consistent applications correct","authors":"Mahsa Najafzadeh, Alexey Gotsman, Hongseok Yang, Carla Ferreira, M. Shapiro","doi":"10.1145/2911151.2911160","DOIUrl":"https://doi.org/10.1145/2911151.2911160","url":null,"abstract":"Designers of a replicated database face a vexing choice between strong consistency, which ensures certain application invariants but is slow and fragile, and asynchronous replication, which is highly available and responsive, but exposes the programmer to unfamiliar behaviours. To bypass this conundrum, recent research has studied hybrid consistency models, in which updates are asynchronous by default, but synchronisation is available upon request. To help programmers exploit hybrid consistency, we propose the first static analysis tool for proving integrity invariants of applications using databases with hybrid consistency models. This allows a programmer to find minimal consistency guarantees sufficient for application correctness.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122979081","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}
Most of the database products developed by Basho Technologies are eventually consistent systems, such as Riak KV and Riak Time Series [4]. If Basho were to apply its expertise and experience in eventual consistency systems to a new file store product, then how would such a system be managed?
Basho Technologies开发的大多数数据库产品最终都是一致的系统，例如Riak KV和Riak Time Series[4]。如果Basho将其在最终一致性系统方面的专业知识和经验应用于新的文件存储产品，那么如何管理这样的系统?
{"title":"Coordinating distributed system configuration changes with humming consensus","authors":"S. Fritchie","doi":"10.1145/2911151.2911165","DOIUrl":"https://doi.org/10.1145/2911151.2911165","url":null,"abstract":"Most of the database products developed by Basho Technologies are eventually consistent systems, such as Riak KV and Riak Time Series [4]. If Basho were to apply its expertise and experience in eventual consistency systems to a new file store product, then how would such a system be managed?","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132272450","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}
Christian Weilbach, Konrad Kühne, Annette Bieniusa
We have designed a new kind of confluent replicated datatype that resembles a distributed version control system. It retains the order of write operations at the configurable trade-off of reduced availability or after-the-fact conflict resolution through users or other coordination strategies. We discuss these strategies and explain the concept with an example. Together with our replication middleware the datatype allows to build applications with distributed state in a similar fashion as applications with exclusive local state, while transparently exposing the necessary compromises in terms of the CAP theorem.
{"title":"Decoupling conflict resolution with CDVCS","authors":"Christian Weilbach, Konrad Kühne, Annette Bieniusa","doi":"10.1145/2911151.2911154","DOIUrl":"https://doi.org/10.1145/2911151.2911154","url":null,"abstract":"We have designed a new kind of confluent replicated datatype that resembles a distributed version control system. It retains the order of write operations at the configurable trade-off of reduced availability or after-the-fact conflict resolution through users or other coordination strategies. We discuss these strategies and explain the concept with an example. Together with our replication middleware the datatype allows to build applications with distributed state in a similar fashion as applications with exclusive local state, while transparently exposing the necessary compromises in terms of the CAP theorem.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132447493","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 a novel approach to the verification of consistency models implemented in distributed storage systems. We base our work on a declarative semantic model defining consistency conditions as predicates expressing ordering and visibility of operations. This model allows for a testing methodology focused on correctness properties rather than operational semantics. Finally, we present and discuss the design and preliminary implementation of a practical property-based consistency verification framework.
{"title":"Towards property-based consistency verification","authors":"P. Viotti, Christopher S. Meiklejohn, M. Vukolic","doi":"10.1145/2911151.2911162","DOIUrl":"https://doi.org/10.1145/2911151.2911162","url":null,"abstract":"We propose a novel approach to the verification of consistency models implemented in distributed storage systems. We base our work on a declarative semantic model defining consistency conditions as predicates expressing ordering and visibility of operations. This model allows for a testing methodology focused on correctness properties rather than operational semantics. Finally, we present and discuss the design and preliminary implementation of a practical property-based consistency verification framework.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131403932","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}
Non-Monotonic Snapshot Isolation (NMSI), a variant of the widely deployed Snapshot Isolation (SI), aims at improving scalability by relaxing snapshots. In contrast to SI, NMSI snapshots are causally consistent, which allows for more parallelism and a reduced abort rate. This work documents the design of PhysiCS-NMSI, a transactional protocol implementing NMSI in a partitioned data store. It is the first protocol to rely on a single scalar taken from a physical clock for tracking causal dependencies and building causally consistent snapshots. Its commit protocol ensures atomicity and the absence of write-write conflicts. We argue that PhysiCS-NMSI approach increases concurrency and reduces abort rate and metadata overhead as compared to state-of-art systems.
{"title":"PhysiCS-NMSI: efficient consistent snapshots for scalable snapshot isolation","authors":"Alejandro Z. Tomsic, Tyler Crain, M. Shapiro","doi":"10.1145/2911151.2911166","DOIUrl":"https://doi.org/10.1145/2911151.2911166","url":null,"abstract":"Non-Monotonic Snapshot Isolation (NMSI), a variant of the widely deployed Snapshot Isolation (SI), aims at improving scalability by relaxing snapshots. In contrast to SI, NMSI snapshots are causally consistent, which allows for more parallelism and a reduced abort rate. This work documents the design of PhysiCS-NMSI, a transactional protocol implementing NMSI in a partitioned data store. It is the first protocol to rely on a single scalar taken from a physical clock for tracking causal dependencies and building causally consistent snapshots. Its commit protocol ensures atomicity and the absence of write-write conflicts. We argue that PhysiCS-NMSI approach increases concurrency and reduces abort rate and metadata overhead as compared to state-of-art systems.","PeriodicalId":259835,"journal":{"name":"Proceedings of the 2nd Workshop on the Principles and Practice of Consistency for Distributed Data","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130042640","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：