How consistent is your cloud application?

Abstract

Current cloud datastores usually trade consistency for performance and availability. However, it is often not clear how an application is affected when it runs under a low level of consistency. In fact, current application designers have basically no tools that would help them to get a feeling of which and how many inconsistencies actually occur for their particular application. In this paper, we propose a generalized approach for detecting consistency anomalies for arbitrary cloud applications accessing various types of cloud datastores in transactional or non-transactional contexts. We do not require any knowledge on the business logic of the studied application nor on its selected consistency guarantees. We experimentally verify the effectiveness of our approach by using the Google App Engine and Cassandra datastores.

Benchmarking Web API Quality

“…For instance, the de facto standard YCSB [29] and its extensions [30], [31] introduced database benchmarking based on CRUD interfaces, which are more compatible with modern NoSQL stores like Apache Cassandra [32]. Other approaches such as OLTPBench [33] or BenchFoundry [34], [35] aim to build comprehensive multi-quality benchmarking platforms that also include measurement approaches for qualities beyond performance, e.g., data consistency [10], [36]- [39] or elastic scalability [11], [12]. Beyond this, there are a number of approaches studying performance impacts of TLS on NoSQL datastores [40]- [42], web services [43], and web servers [44].…”

Feral Concurrency Control

“…Wada et al evaluated the staleness of Amazon's SimpleDB using end-user request tracing [81], while Bermbach and Tai evaluated Amazon S3 [22], each quantifying various forms of non-serializable behavior. Golab et al provide algorithms for verifying the linearizability of and sequential consistency arbitrary data stores [51] and Zellag and Kemme provide algorithms for verifying their serializability [85] and other cycle-based isolation anomalies [86]. Probabilistically Bounded Staleness provides time-and version-based staleness predictions for eventually consistent data stores [18].…”

Benchmarking Eventual Consistency: Lessons Learned from Long-Term Experimental Studies

“…Zellag and Kemme [31] also present an alternative approach counting consistency anomalies for arbitrary cloudbased applications in transactional and non-transactional datastores. At runtime, their approach builds a dependency graph to detect cycles, i.e., consistency violations.…”