Databases (cs.DB)

Abstract: Transaction processing is of growing importance for mobile and web applications. Booking tickets, flight reservation, e-Banking, e-Payment, and booking holiday arrangements are just a few examples. Due to temporarily disconnected situations the synchronization and consistent transaction processing are key issues. To avoid difficulties with blocked transactions or communication loss several authors and technology providers have recommended to use Optimistic Concurrency Control (OCC) to solve the problem. However most vendors of Relational Database Management Systems (DBMS) implemented only locking schemes for concurrency control which prohibit the immediate use of OCC. We propose Row Version Verifying (RVV) discipline to avoid lost updates and achieve a kind of OCC for those DBMS not providing an adequate non-blocking concurrency control. Moreover, the different mechanisms are categorized as access pattern in order to provide programmers with a general guideline for SQL databases. The proposed SQL access patterns are relevant for all transactional applications with unreliable communication and low conflicting situations. We demonstrate the proposed solution using mainstream database systems like Oracle, DB2, and SQLServer.

Abstract: This paper presents a concurrency control mechanism that does not follow a 'one concurrency control mechanism fits all needs' strategy. With the presented mechanism a transaction runs under several concurrency control mechanisms and the appropriate one is chosen based on the accessed data. For this purpose, the data is divided into four classes based on its access type and usage (semantics). Class $O$ (the optimistic class) implements a first-committer-wins strategy, class $R$ (the reconciliation class) implements a first-n-committers-win strategy, class $P$ (the pessimistic class) implements a first-reader-wins strategy, and class $E$ (the escrow class) implements a first-n-readers-win strategy. Accordingly, the model is called \PeFS. The selected concurrency control mechanism may be automatically adapted at run-time according to the current load or a known usage profile. This run-time adaptation allows \Pe to balance the commit rate and the response time even under changing conditions. \Pe outperforms the Snapshot Isolation concurrency control in terms of response time by a factor of approximately 4.5 under heavy transactional load (4000 concurrent transactions). As consequence, the degree of concurrency is 3.2 times higher.

Abstract: Analysts and scientists are interested in querying streams of video, audio, and text to extract quantitative insights. For example, an urban planner may wish to measure congestion by querying the live feed from a traffic camera. Prior work has used deep neural networks (DNNs) to answer such queries in the batch setting. However, much of this work is not suited for the streaming setting because it requires access to the entire dataset before a query can be submitted or is specific to video. Thus, to the best of our knowledge, no prior work addresses the problem of efficiently answering queries over multiple modalities of streams. In this work we propose InQuest, a system for accelerating aggregation queries on unstructured streams of data with statistical guarantees on query accuracy. InQuest leverages inexpensive approximation models ("proxies") and sampling techniques to limit the execution of an expensive high-precision model (an "oracle") to a subset of the stream. It then uses the oracle predictions to compute an approximate query answer in real-time. We theoretically analyzed InQuest and show that the expected error of its query estimates converges on stationary streams at a rate inversely proportional to the oracle budget. We evaluated our algorithm on six real-world video and text datasets and show that InQuest achieves the same root mean squared error (RMSE) as two streaming baselines with up to 5.0x fewer oracle invocations. We further show that InQuest can achieve up to 1.9x lower RMSE at a fixed number of oracle invocations than a state-of-the-art batch setting algorithm.