Verifying eventual consistency of optimistic replication systems

Bouajjani, Ahmed; Enea, Constantin; Hamza, Jad

doi:10.1145/2535838.2535877

Cited by 38 publications

(36 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, it was shown that checking linearizability is an EXPSPACE-complete problem when the number of sites is bounded [3,21]. Eventual consistency has been shown to be decidable [8]. Sequential consistency was shown to be undecidable [3].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

On verifying causal consistency

Bouajjani

Enea

Guerraoui

et al. 2017

Proceedings of the 44th ACM SIGPLAN Symposium on Principles of Programming Languages

Self Cite

View full text Add to dashboard Cite

Abstract. Causal consistency is one of the most adopted consistency criteria for distributed implementations of data structures. It ensures that operations are executed at all sites according to their causal precedence. We address the issue of verifying automatically whether the executions of an implementation of a data structure are causally consistent. We consider two problems: (1) checking whether one single execution is causally consistent, which is relevant for developing testing and bug finding algorithms, and (2) verifying whether all the executions of an implementation are causally consistent.We show that the first problem is NP-complete. This holds even for the read-write memory abstraction, which is a building block of many modern distributed systems. Indeed, such systems often store data in key-value stores, which are instances of the readwrite memory abstraction. Moreover, we prove that, surprisingly, the second problem is undecidable, and again this holds even for the read-write memory abstraction. However, we show that for the read-write memory abstraction, these negative results can be circumvented if the implementations are data independent, i.e., their behaviors do not depend on the data values that are written or read at each moment, which is a realistic assumption.We prove that for data independent implementations, the problem of checking the correctness of a single execution w.r.t. the read-write memory abstraction is polynomial time. Furthermore, we show that for such implementations the set of non-causally consistent executions can be represented by means of a finite number of register automata. Using these machines as observers (in parallel with the implementation) allows to reduce polynomially the problem of checking causal consistency to a state reachability problem. This reduction holds regardless of the class of programs used for the implementation, of the number of read-write variables, and of the used data domain. It allows leveraging existing techniques for assertion/reachability checking to causal consistency verification. Moreover, for a significant class of implementations, we derive from this reduction the decidability of verifying causal consistency w.r.t. the read-write memory abstraction.

show abstract

Section: Related Workmentioning

confidence: 99%

“…This undecidability result might be surprising, since it is known that linearizability (stronger than CC) [23] and eventual consistency (weaker than CC) [39] are decidable to verify in that same setting [3,8,21]. This result reveals an interesting aspect in the definition of causal consistency.…”

Section: Introductionmentioning

confidence: 95%

On verifying causal consistency

Bouajjani

Enea

Guerraoui

et al. 2017

Proceedings of the 44th ACM SIGPLAN Symposium on Principles of Programming Languages

Self Cite

View full text Add to dashboard Cite

show abstract

“…In [14,15], for instance, the focus was on abstract properties, while in [2] the goal was to show how to verify replication schemes that ensure only eventual consistency. However, little effort has been devoted to theoretical studies that would precisely measure the power and limitations of eventual consistency (beyond artifacts related to specific implementations).…”

Section: Related Workmentioning

confidence: 99%

“…This is true even if we only require the resulting linearizable object to be obstruction-free, whereas the eventually linearizable base objects are waitfree. (2) No set of eventually linearizable objects can boost the consensus power of simple linearizable objects like registers, i.e., help solve consensus among two processes in conjunction with linearizable registers. In short, eventually linearizable objects can neither implement linearizable ones (1) nor augment the power of shared linearizable registers (2).…”

Section: Introductionmentioning

confidence: 99%

A paradox of eventual linearizability in shared memory

Guerraoui

Ruppert

2014

Proceedings of the 2014 ACM Symposium on Principles of Distributed Computing

View full text Add to dashboard Cite

This paper compares, for the first time, the computational power of linearizable objects with that of eventually linearizable ones. We present the following paradox. We show that, unsurprisingly, no set of eventually linearizable objects can (1) implement any non-trivial linearizable object, nor (2) boost the consensus power of simple objects like linearizable registers. We also show, perhaps surprisingly, that any implementation of an eventually linearizable complex object like a fetch&increment counter (from linearizable base objects), can itself be viewed as a fully linearizable implementation of the same fetch&increment counter (using the exact same set of base objects).

show abstract

“…[10] specifies replicated data types using relations over events and proposes a framework for verifying store implementations. On another line of research, the work in [7] focuses on criteria for decidable checking of eventual consistency of systems.…”

Section: Related Workmentioning

confidence: 99%

On the behaviour of general purpose applications on cloud storages

Bocchi

Melgratti

2014

SOCA

View full text Add to dashboard Cite

Managing data over cloud infrastructures raises novel challenges with respect to existing and well studied approaches such as ACID and long running transactions. One of the main requirements is to provide availability and partition tolerance in a scenario with replicas and distributed control. This comes at the price of a weaker consistency, usually called eventual consistency. These weak memory models have proved to be suitable in a number of scenarios, such as the analysis of large data with map-reduce. However, due to the widespread availability of cloud infrastructures, weak storages are used not only by specialised applications but also by general purpose applications. We provide a formal approach, based on process calculi, to reason about the behaviour of programs that rely on cloud stores. For instance, it allows to check that the composition of a process with a cloud store ensures 'strong' properties through a wise usage of asynchronous message-passing; in this case we say that the process supports the consistency level provided by the cloud store. The proposed approach is compositional: the support of a consistency level is preserved by parallel composition when the preorder used to compare process-store ensembles is the weak simulation.

show abstract

Verifying eventual consistency of optimistic replication systems

Cited by 38 publications

References 23 publications

On verifying causal consistency

On verifying causal consistency

A paradox of eventual linearizability in shared memory

On the behaviour of general purpose applications on cloud storages

Contact Info

Product

Resources

About