We study efficient and robust implementations of an atomic read-write data structure over an asynchronous distributed message-passing system made of reader and writer processes, as well as a number of servers implementing the data structure. We determine the exact conditions under which every read and write involves one round of communication with the servers. These conditions relate the number of readers to the tolerated number of faulty servers and the nature of these failures.

Total order broadcast is a fundamental communication primitive that plays a central role in bringing cheap software-based high availability to a wide range of services. This article studies the practical performance of such a primitive on a cluster of homogeneous machines. We present LCR, the first throughput optimal uniform total order broadcast protocol. LCR is based on a ring topology. It only relies on point-to-point inter-process communication and has a linear latency with respect to the number of processes. LCR is also fair in the sense that each process has an equal opportunity of having its messages delivered by all processes. We benchmark a C implementation of LCR against Spread and JGroups, two of the most widely used group communication packages. LCR provides higher throughput than the alternatives, over a large number of scenarios.

This article considers the problem of robustly emulating a shared atomic memory over a distributed message-passing system where processes can fail by crashing and possibly recover. We revisit the notion of atomicity in the crash-recovery context and introduce a generic algorithm that emulates an atomic memory. The algorithm is instantiated for various settings according to whether processes have access to local stable storage, and whether, in every execution of the algorithm, a sufficient number of processes are assumed not to crash. We establish the optimality of specific instances of our algorithm in terms of resilience , log complexity (number of stable storage accesses needed in every read or write operation), as well as time complexity (number of communication steps needed in every read or write operation). The article also discusses the impact of considering a multiwriter versus a single-writer memory, as well as the impact of weakening the consistency of the memory by providing safe or regular semantics instead of atomicity.

This paper presents an algorithm to ensure the atomicity of a distributed storage that can be read and written by any number of clients. In failure-free and synchronous situations, and even if there is contention, our algorithm has a high write throughput and a read throughput that grows linearly with the number of available servers. The algorithm is devised with a homogeneous cluster of servers in mind. It organizes servers around a ring and assumes point-to-point communication. It is resilient to the crash failure of any number of readers and writers as well as to the crash failure of all but one server. We evaluated our algorithm on a cluster of 24 nodes with dual fast ethernet network interfaces (100 Mbps). We achieve 81 Mbps of write throughput and 8×90 Mbps of read throughput (with up to 8 servers) which conveys the linear scalability with the number of servers.

Total order broadcast is a fundamental communication primitive that plays a central role in bringing cheap software-based high availability to a wide array of services. This paper studies the practical performance of such a primitive on a cluster of homogeneous machines. We present FSR, a (uniform) total order broadcast protocol that provides high throughput, regardless of message broadcast patterns. FSR is based on a ring topology, only relies on point-to-point inter-process communication, and has a linear latency with respect to the total number of processes in the system. Moreover, it is fair in the sense that each process has an equal opportunity of having its messages delivered by all processes. On a cluster of Itanium based machines, FSR achieves a throughput of 79 Mbit/s on a 100 Mbit/s switched Ethernet network

This paper establishes tight bounds on the best-case time-complexity of distributed atomic read/write storage implementations that tolerate worst-case conditions. We study asynchronous robust implementations where a writer and a set of reader processes (clients) access an atomic storage implemented over a set of 2t+b+1 server processes of which t can fail: b of these can be malicious and the rest can crash. We define a lucky operation (read or write) as one that runs synchronously and without contention. It is often argued in practice that lucky operations are the most frequent. We determine the exact conditions under which a lucky operation can be fast, namely expedited in onecommunication round-trip with no data authentication. We show that every lucky write (resp., read) can be fast despite fw(resp., fr) actual failures, if and only if fw + fr \lt t-b.

A shared memory abstraction can be robustly emulated over an asynchronous message passing system where any process can fail by crashing and possibly recover (crash-recovery model), by having (a) the processes exchange messages to synchronize their read and write operations and (b) log key information on their local stable storage. This paper extends the existing atomicity consistency criterion defined for multiwriter/multireader shared memory in a crash-stop model, by providing two new criteria for the crash-recovery model. We introduce lower bounds on the log-complexity for each of the two corresponding types of robust shared memory emulations. We demonstrate that our lower bounds are tight by providing algorithms that match them. Besides being optimal, these algorithms have the same message and time complexity as their most efficient counterpart we know of in the crash-stop model.

This paper addresses the problem of designing an efficient implementation of a basic atomic read-write data structure over an asynchronous message-passing system. In particular, we consider time-efficient implementations of this abstraction in the case of a single writer, multiple readers (also called a SWMR atomic register) and S servers: the writer, the readers, and t out of the S servers may fail by crashing. Previous implementations tolerate the failure of any minority of servers (i.e., t < S/2) and require one communication round-trip for every write, and two round-trips for every read.We investigate the possibility of fast implementations, namely, implementations that complete both reads and writes in one round-trip. We show that, interestingly, the existence of a fast implementation depends on the maximum number of readers considered. More precisely, we show that a fast implementation is possible if and only if the number of readers is less that S t-2. We also show that a fast implementation is impossible in a multiple writers setting when t ≥ 1.Our results draw sharp lines between the time-complexity of regular and atomic register implementations, as well as between single-writer and multi-writer implementations. The results lead also to revisit, in a message-passing context, the folklore theorem that "atomic reads must write".