143 research outputs found
OASIcs, Volume 101, FAB 2022, Complete Volume
No full textOASIcs, Volume 101, FAB 2022, Complete Volum
Front Matter, Table of Contents, Preface, Conference Organization
No full textFront Matter, Table of Contents, Preface, Conference Organizatio
Granular Synchrony
No full textToday’s mainstream network timing models for distributed computing are synchrony, partial synchrony, and asynchrony. These models are coarse-grained and often make either too strong or too weak assumptions about the network. This paper introduces a new timing model called granular synchrony that models the network as a mixture of synchronous, partially synchronous, and asynchronous communication links. The new model is not only theoretically interesting but also more representative of real-world networks. It also serves as a unifying framework where current mainstream models are its special cases. We present necessary and sufficient conditions for solving crash and Byzantine fault-tolerant consensus in granular synchrony. Interestingly, consensus among n parties can be achieved against f ≥ n/2 crash faults or f ≥ n/3 Byzantine faults without resorting to full synchrony
Brief Announcement: It’s not easy to relax: liveness in chained BFT protocols
No full textModern chained BFT SMR protocols have poor liveness under failures as they require multiple consecutive honest leaders to commit a single block. Siesta, our proposed new BFT SMR protocol, is instead able to commit a block that spans multiple non-consecutive honest leaders. Siesta reduces the expected commit latency of HotStuff by a factor of three under failures, and the worst-case latency by a factor of eight
Recommended from our members
A client-centric approach to transactional datastores
No full textModern applications must collect and store massive amounts of data. Cloud storage offers these applications simplicity: the abstraction of a failure-free, perfectly scalable black-box. While appealing, offloading data to the cloud is not without its challenges. These cloud storage systems often favour weaker levels of isolation and consistency. These weaker guarantees introduce behaviours that, without care, can break application logic. Offloading data to an untrusted third party like the cloud also raises questions of security and privacy. This thesis seeks to improve the performance, the semantics and the security of transactional cloud storage systems. It centers around a simple idea: defining consistency guarantees from the perspective of the applications that observe these guarantees, rather than from the perspective of the systems that implement them. This new perspective brings forth several benefits. First, it offers simpler and cleaner definitions of weak isolation and consistency guarantees. Second, it enables scalable implementations of existing guarantees like causal consistency. Finally, it has applications to security: it allows us to efficienctly augment transactional cloud storage systems with obliviousness guaranteesComputer Scienc
Recommended from our members
Minimizing I/O bottlenecks to achieve scalable and high-throughput systems
No full textModern applications store large volumes of data and access this data at high throughput. Fortunately, advanced hardware meets applications’ demands by offering low-latency and high-bandwidth storage and datacenter networks. However, state-of-the-art systems infrastructures underutilize available hardware resources and fail to meet the throughput and scalability requirements of I/O-intensive applications. This dissertation studies the performance limitations of three distinct systems: monolithic key-value stores, distributed transactional stores, and public blockchains. First, it attributes their low throughput and poor scalability to I/O-bottlenecks that are inherent to the systems’ design and architecture. Next, it addresses the question: How do we architect systems to minimize I/O-bottlenecks and simultaneously achieve high throughput and scalability? It proposes a fundamental redesign of systems by carefully crafting the roles and responsibilities of each system component to improve the utilization of underlying resources. This dissertation employs a few key ideas to achieve scalable, high-throughput systems. The first idea aims at customizing the storage subsystem to align with the performance characteristics of underlying media and its applications’ data layout. The second centers on co-designing data processing along with its storage with a clear demarcation of the roles of each system component. The third focuses on leveraging asynchrony and batching to reconstruct how and when I/O is performed for improving the utilization of available resources. This dissertation revisits these well-known ideas, however, from the perspective of minimizing I/O-bottlenecks to achieve high throughput and scalability. This dissertation combines these ideas to architect three novel systems; Skye: a monolithic key-value store, Cascades: a distributed transactional store, and RainBlock: a distributed and decentralized database. Skye is tailored for Persistent Memory (PM), a novel storage-class memory technology that offers high throughput and low latency. Cascades is optimized for cloud servers and SSDs interconnected through high-speed datacenter networks. Finally, RainBlock is designed for commodity hardware and targets public environments with untrusted servers. This work presents the design and architecture of each of these systems, discusses their trade-offs, and evaluates their end-to-end performance and scalability.Computer Scienc
Kedrovyi Sor : la vie quotidienne dans un camp du goulag à l'époque stalinienne
No full textKedrovyi Sor: daily life in a Gulag camp during the Stalinist period, Oleg Azarov.
Our knowledge of the Soviet concentration camp universe is beginning to be nourished by direct access to archives. The author, a young Russian scholar, has been able to work on a rich collection of documents concerning the camp of Kedrovyi Sor, in the Pečora bassin. Through his study of the camp's administrative structure, its financial operation and its economic activities, and through an analysis of the inmates and their living conditions, he provides, almost in the raw, a concrete and precise description of the daily functioning of a Stalinist camp from the early 1930s to the 1950s.Azarov Oleg, Laurent Natacha. Kedrovyi Sor : la vie quotidienne dans un camp du goulag à l'époque stalinienne. In: Vingtième Siècle, revue d'histoire, n°43, juillet-septembre 1994. Dossier : Histoire au présent de la "political correctness" pp. 69-87
- …
