Dotted Version Vectors: Logical Clocks for Optimistic Replication

2005

Abstract Data replication is a key technology in distributed systems that enables higher availability and performance. This article surveys optimistic replication algorithms. They allow replica contents to diverge in the short term to support concurrent work practices and tolerate failures in low-quality communication links. The importance of such techniques is increasing as collaboration through wide-area and mobile networks becomes popular.

Concurrency and Computation: Practice and Experience, 2018

Several distributed services ranging from key-value stores to cloud storage require fault-tolerance and reliability features. For enabling fast recovery and seamless transition, primary-backup replication protocols are widely used in different application settings including distributed databases, web services, and the Internet of Things. In this study, we elaborate the ways of enhancing the efficiency of the primary-backup replication protocol by introducing various checkpointing techniques. We develop a geographically replicated key-value store based on the RocksDB and use the PlanetLab testbed network for large-scale performance analysis. Using various metrics of interest including blocking time, checkpointing time, checkpoint size, failover time, and throughput and testing with practical workloads via the YCSB tool, our findings indicate that periodic-incremental checkpointing promises up to 5 times decrease in blocking time and a drastic improvement on the overall throughput compared to the traditional primary-backup replication. Furthermore, enabling Snappy compression algorithm on the periodic-incremental checkpointing leads to further reduction in blocking time and increases system throughput compared to the traditional primary-backup replication. KEYWORDS checkpointing, compressed checkpointing, incremental checkpointing, periodic checkpointing, primary-backup replication, replicated cloud key-value stores 1 INTRODUCTION As the cloud systems continue to enlarge, the underlying networks empowering them also maintain their steady growth to stay sustainable against challenges involving immense user population and the big data. This growth is observed in two aspects as the geographical scaling of the nodes and the increase in the node counts. The availability becomes more and more significant as any outages that could last milliseconds of increase in response times may result in high income losses. 1 Moreover, the possibility of facing with failures in these systems is inevitable due to extensive usage of software and hardware components with the long running applications that exceed the mean time between failures of the components. 2 The most important and effective approach to deal with crash failures is replication. It is widely used as a fault-tolerance mechanism, and finding optimal replication protocols is an active research area. There exist two main types of replication protocols, namely, active and passive. In the active replication, which is also known as state-machine replication, every incoming request is processed by every replica in the system resulting in multiple results to be collected. Once collected, they are reduced into a single result value using various algorithms and the client is notified accordingly. In the passive replication, which is also known as primary-backup replication, there exist a single primary replica and a group of backup replicas. Each request is executed only in the primary replica, the result is then copied to backup replicas and the client is notified. Another way of introducing recovery from failures is through the checkpointing that refers to saving the system state to a stable storage after critical executions. Afterwards, in the event of any failures during the execution, the previously saved checkpoint can be restored as a failure-free system state enabling the execution continue over. This approach also facilitates a quick rollback feature even against unforeseen failures and decreases the workload needed to revitalize a replica from zero state, since with a single rollback, the system state would be caught up with the latest failure-free state. 3 In our recent work, we demonstrated applicability and benefits of various checkpointing algorithms in replication protocols. 4,5

IEEE/IFIP International Conference on Dependable Systems and Networks (DSN 2012), 2012

Deferred update replication is a well-known approach to building data management systems as it provides both high availability and high performance. High availability comes from the fact that any replica can execute client transactions; the crash of one or more replicas does not interrupt the system. High performance comes from the fact that only one replica executes a transaction; the others must only apply its updates. Since replicas execute transactions concurrently, transaction execution is distributed across the system. The main drawback of deferred update replication is that update transactions scale poorly with the number of replicas, although read-only transactions scale well. This paper proposes an extension to the technique that improves the scalability of update transactions. In addition to presenting a novel protocol, we detail its implementation and provide an extensive analysis of its performance.

Journal of Internet Services and Applications

Cloud computing is a general term that involves delivering hosted services over the Internet. With the accelerated growth of the volume of data used by applications, many organizations have moved their data into cloud servers to provide scalable, reliable and highly available services. A particularly challenging issue that arises in the context of cloud storage systems with geographically-distributed data replication is how to reach a consistent state for all replicas. This survey reviews major aspects related to consistency issues in cloud data storage systems, categorizing recently proposed methods into three categories: (1) fixed consistency methods, (2) configurable consistency methods and (3) consistency monitoring methods.

1994

This research was sponsored by the Air Force Materiel Command (AFMC) and the Advanced Research Projects Agency (ARPA) under contract number F19628-93-C-0193. Additional support was provided by the IBM Corporation, Digital Equipment Corporation, Bellcore, and Intel ...

Data replication is a common technique used for fault-tolerance in reliable distributed systems. In geo-replicated systems and the cloud, it additionally provides low latency. Recently, causal consistency in such systems has received much attention. However, all existing works assume the data is fully replicated. This greatly simplifies the design of the algorithms to implement causal consistency. In this paper, we propose that it can be advantageous to have partial replication of data, and we propose two algorithms for achieving causal consistency in such systems where the data is only partially replicated. This is the first work that explores causal consistency for partially replicated geo-replicated systems. We also give a special case algorithm for causal consistency in the full-replication case.

2019 IEEE 18th International Symposium on Network Computing and Applications (NCA), 2019

Data management has become crucial. Distributed applications and users manipulate large amounts of data. More and more distributed data management solutions arise, e.g. Casandra or Cosmos DB. Some of them propose multiple consistency protocols. Thus, for each piece of data, the developer or the user can choose a consistency protocol adapted to his needs. In this paper we explain why taking the consistency protocol into account is important while replication (especially placing) pieces of data; and we propose CAnDoR, an approach that dynamically adapts the replication according to the data usage (read/write frequencies and locations) and the consistency protocol used to manage the piece of data. Our simulations show that using CAnDoR to place and move data copies can improve the global average access latency by up to 40%.

2018

The problem of ensuring consistency in applications that manage replicated data is one of the main challenges of distributed computing. Among the several invariants that may be enforced, ensuring that updates are applied and made visible respecting causality has emerged as a key ingredient among the many consistency criteria and client session guarantees that have been proposed and implemented in the last decade. Techniques to keep track of causal dependencies, and to subsequently ensure that messages are delivered in causal order, have been widely studied. It is today well known that, in order to accurately capture causality one may need to keep a large amounts of metadata, for instance, one vector clock for each data object. This metadata needs to be updated and piggybacked on update messages, such that updates that are received from remote datacenters can be applied locally without violating causality. This metadata can be compressed; ultimately, it is possible to preserve causal...

Information Processing Letters, 2010

log version vector Replication system In a replication system, version vectors are logged with replicas to detect conflicts among operations. Dynamic replications where replicas are frequently created and destroyed suffer from expensive logging overhead caused by inactive entries of version vectors. Although the rigmarole of pruning vectors can delete inactive entries, the vectors may be incompatible without additional information, which also causes another overhead. This paper proposes a novel version vector called log (log-prime) consisting of only three entries. By encoding based on the characteristics of prime numbers, log version vectors of fixed size can be logged concisely with no pruning technique at a little sacrifice in accuracy. Simulation studies show that log version vectors are accurate enough to detect almost all conflicts in the replication systems where all replicas are fully synchronizing.

ACM SIGMOD Record, 1999

Replication is often used in many distributed systems to provide a higher level of performance, reliability and availability. Lazy replica update protocols, which propagate updates to replicas through independent transactions after the original transaction commits, have become popular with database vendors due to their superior performance characteristics. However, if lazy protocols are used indiscriminately, they can result in non-serializable executions. In this paper, we propose two new lazy update protocols that guarantee serializability but impose a much weaker requirement on data placement than earlier protocols. Further, many naturally occurring distributed systems, like distributed data warehouses, satisfy this requirement. We also extend our lazy update protocols to eliminate all requirements on data placement. The extension is a hybrid protocol that propagates as many updates as possible in a lazy fashion. We implemented our protocols on the Datablitz database system product developed at Bell Labs. We also conducted an extensive performance study which shows that our protocols outperform existing protocols over a wide range of workloads.