System support for object groups

Guerraoui, Rachid; Felber, Pascal; Garbinato, Benoît; Mazouni, K.

doi:10.1145/286936.286961

Cited by 32 publications

(14 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In such a context the replicated invocation problem, originally analyzed in [47] and later addressed in, e.g., [26,30,77], shows similarities with the problems we have tackled with our proposal. Specifically, the replicated invocation problem arises when an actively replicated server, say S, invokes, in its turn another (replicated) server, say R. In such a scenario, the processing of the client request at the different S server replicas originates duplicate invocations towards R. This leads to inconsistencies of the state of R if the invoked application logic is not idempotent.…”

Section: Related Workmentioning

confidence: 66%

Design and Evaluation of a Parallel Invocation Protocol for Transactional Applications over the Web

Romano

Quaglia

2014

IEEE Trans. Comput.

View full text Add to dashboard Cite

Edge computing is a powerful tool to face the challenging performance requirements of modern Internet applications. By replicating applications' data and logic across a large number of geographically distributed servers, edge computing platforms allow to achieve significant enhancements of the proximity between clients and contents, and of the system scalability. These platforms reveal highly effective when handling requests entailing readonly access to the application data, as these requests can be autonomously served by some edge server typically located closer to the client than the origin site. However, in contexts where end-users can trigger transactional manipulations of the application state (e.g., e-Commerce, auctions or financial applications), the corresponding update requests typically need to be re-directed to the origin transactional data sources, thus nullifying any performance benefit arising from data replication and client proximity.In order to cope with this issue, in this article we present a parallel invocation protocol, which exploits the path-diversity along the end-to-end interaction towards the origin sites by concurrently routing transactional requests towards multiple edge servers. Request processing is finally carried out by a single edge server, adaptively selected as the most responsive one depending on current system conditions. The proposed edge server selection scheme does not require coordination among (geographically distributed) edge server instances, thus being very light and scalable. The benefits from our protocol in terms of both reduced and more predictable end-to-end latency are quantified via an extended simulation study.

show abstract

Section: Related Workmentioning

confidence: 66%

Design and Evaluation of a Parallel Invocation Protocol for Transactional Applications over the Web

Romano

Quaglia

2014

IEEE Trans. Comput.

View full text Add to dashboard Cite

show abstract

“…They state that "A transactional mechanism should however be integrated within group communication to support multi-server request atomicity." [10]. Jean-Pier Briot proposed Actalk [11], an environment where Actors communicate concurrently with asynchronous message passing.…”

Section: The Need For Transactionsmentioning

confidence: 99%

Transactional memory for smalltalk

Renggli

Nierstrasz

2007

Proceedings of the 2007 International Conference on Dynamic Languages: In Conjunction With the 15th International Smalltalk Joi

View full text Add to dashboard Cite

Abstract. Concurrency control in Smalltalk is based on locks and is therefore notoriously difficult to use. Even though some implementations provide high-level constructs, these add complexity and potentially hard-to-detect bugs to the application. Transactional memory is an attractive mechanism that does not have the drawbacks of locks, however the underlying implementation is often difficult to integrate into an existing language. In this paper we show how we have introduced transactional semantics in Smalltalk by using the reflective facilities of the language. Our approach is based on method annotations, incremental parse tree transformations and an optimistic commit protocol. We report on a practical case study, benchmarks and further and on-going work. Only a few current Smalltalk implementations provide more sophisticated synchronization support. VisualWorks Smalltalk provides a reentrant lock implementation that enables a process to enter the same critical section multiple times. Other processes get blocked until the owning process leaves the critical section. Unfortunately this kind of synchronization has its drawbacks and is notoriously difficult to use [3]:Deadlocks. If there are cyclic dependencies between resources and processes, applications may deadlock. This problem can be avoided by acquiring resources in a fixed order, however in practice this is often difficult to achieve. Starvation. A process that never leaves a critical section, due to a bug in the software or an unforeseen error, will continue to hold the lock forever. Other processes that would like to enter the critical section starve.

show abstract

“…Trading replica consistency for increased availability has been addressed in distributed object systems as [22,23,24]. However, these systems either guarantee strong replica consistency or no replica consistency at all.…”

Section: Related Workmentioning

confidence: 99%