Middleware-based database replication protocols require few or no changes in the database engine. Thus, they are more portable and flexible than kernel-based protocols, but have coarser-grain information about transaction access data, resulting in reduced concurrency and increased aborts. This paper proposes conflict-aware load-balancing techniques to increase the concurrency and reduce the abort rate of middleware-based replication protocols. Our algorithms assign transactions to replicas so that the number of conflicting transactions executing on distinct servers is reduced and the processing load is equitably distributed over the servers. Experimental evaluation using a prototype of our system running the TPC-C benchmark showed that aborts can be reduced with no penalty in response time.

The need for high availability and performance in data management systems has been fueling a long running interest in database replication from both academia and industry. However, academic groups often attack replication problems in isolation, overlooking the need for completeness in their solutions, while commercial teams take a holistic approach that often misses opportunities for fundamental innovation. This has created over time a gap between academic research and industrial practice. This paper aims to characterize the gap along three axes: performance, availability, and administration. We build on our own experience developing and deploying replication systems in commercial and academic settings, as well as on a large body of prior related work. We sift through representative examples from the last decade of open-source, academic, and commercial database replication systems and combine this material with case studies from real systems deployed at Fortune 500 customers. We propose two agendas, one for academic research and one for industrial R&D, which we believe can bridge the gap within 5-10 years. This way, we hope to both motivate and help researchers in making the theory and practice of middleware-based database replication more relevant to each other.

We describe the design and implementation of Walter, a key-value store that supports transactions and replicates data across distant sites. A key feature behind Walter is a new property called Parallel Snapshot Isolation (PSI). PSI allows Walter to replicate data asynchronously, while providing strong guarantees within each site. PSI precludes write-write conflicts, so that developers need not worry about conflict-resolution logic. To prevent write-write conflicts and implement PSI, Walter uses two new and simple techniques: preferred sites and counting sets. We use Walter to build a social networking application and port a Twitter-like application.

The contents of this report are the sole responsibility of the authors. O conteúdo do presente relatório é de única responsabilidade dos autores.

As more and more applications are deployed as Internet-based services, they have to be available anytime anywhere in a seamless manner. This requires the underlying infrastructure to provide scalability, faulttolerance and fast response times. While replicating the services and the data they access across sites that are located in different geographic regions is a promising means to achieve these requirements, data consistency is challenging if data continuously changes and queries are dynamic by nature, as is typical for e-commerce applications. Thus, current WAN replication solutions either trade performance for data consistency or are not able to scale in wide-area settings. In this paper, we present a novel approach to provide performance and consistency for Internet services. One of the main contributions is an autonomic replica placement module that places data copies only on servers close to clients that actually need them. The goal is to find the right trade-off between fast local access and the overhead of keeping data copies consistent. As data access patterns might change over time, reconfiguration is done periodically and online, i.e., allowing sites to receive new data copies or drop data copies while at the same time transaction processing continues in the system. 1.

This paper develops analytical models to predict the throughput and the response time of a replicated database using measurements of the workload on a standalone database. These models allow workload scalability to be estimated before the replicated system is deployed, making the technique useful for capacity planning and dynamic service provisioning. The models capture the scalability limits stemming from update propagation and aborts for both multi-master and single-master replicated databases that support snapshot isolation. We validate the models by comparing their throughput and response time predictions against experimental measurements on two prototype replicated database systems running the TPC-W and RUBiS workloads. We show that the model predictions match the experimental results for both the multi-master and single-master designs and for the various workload mixes of TPC-W and RUBiS.

Recent industry trends towards reducing the costs of ownership in large data centers emphasize the need for database system techniques for both automatic performance tuning and efficient resource usage. The goal is to host several database applications on a shared server farm, including scheduling multiple applications on the same physical server or even within a single database engine, while meeting each application’s service level agreement. Automatic provisioning of database servers to applications and virtualization techniques, such as, live virtual machine migration have been proposed as useful tools to address this problem. In this paper we argue that by allocating entire server boxes and migrating entire application stacks in cases of server overload, these solutions are too coarse-grained for many overload situations. Hence, they may result in resource usage inefficiency, performance penalties, or both. We introduce an outlier detection algorithm which zooms in to the fine-grained query contexts which are most affected by an environment change and/or where a perceived overload problem is likely to originate from. We show that isolating these query contexts through either memory quota enforcements or fine-grained load balancing across different database replicas of their respective applications allows us to alleviate resource interference in many cases of overload. 1

Snapshot isolation has received a considerable amount of attention in the context of full database replication. Such popularity is mainly because read-only transactions executing under snapshot isolation are never blocked or aborted. In partial replication, where each replica holds only a part of the database, transactions may require access to remote databases. Each remote read operation of the transaction must execute in a consistent global database snapshot as the local operations; if such a snapshot is not available, the transaction must be aborted. In this paper we are interested in the effects of distributed transactions on the abort rate of partially replicated snapshot isolation systems. We present a simple probabilistic analysis of transaction abort rates for two different concurrency control mechanisms: lock- and version-based. The former models the behavior of a replication protocol providing one-copy-serializability; the latter models snapshot isolation. Our analysis reveals that in the version-based system the execution abort rate decreases exponentially as the number of data versions available increases. As a consequence, in all cases considered, two versions of each data item were sufficient to eliminate aborts due to distributed transactions. 1.

Databases are fully replicated in order to get two complementary features: performance improvement and high availability. Performance can be improved when a database is replicated since each replica can serve read-only accesses without requiring any coordination with the rest of replicas. Thus, when most of the

This work assesses how crashes and recoveries affect the performance of a replicated dynamic content web application. RobustStore is the result of retrofitting TPC-W’s on-line bookstore with Treplica, a middleware for building dependable applications. Implementations of Paxos and Fast Paxos are at the core of Treplica’s efficient and programmer-friendly support for replication and recovery. The TPC-W benchmark, augmented with faultloads and dependability measures, is used to evaluate the behaviour of RobustStore. Experiments apply faultloads that cause sequential and concurrent replica crashes. RobustStore’s performance drops by less than 13 % during the recovery from two simultaneous replica crashes. When subject to an identical faultload and a shopping workload, a five-replicas RobustStore maintains an accuracy of 99.999%. Our results display not only good performance, total autonomy and uninterrupted availability, they also show that it is simple to develop efficient recovery-oriented applications using Treplica. 1