Quality-Aware Runtime Adaptation in Complex Event Processing

Abstract

Complex event processing (CEP) is a fundamental paradigm for a software system to self-adapt to environmental changes. CEP provides efficient means to detect (complex) events corresponding to environmental changes by performing a realtime analysis on many, possibly heterogeneous, data sources. The way current CEP systems detect events is determined at design time without accounting for dynamic changes of the environment monitored by the CEP system. This can lead to situations where the performance, quality and reliability of event detection significantly drop (e.g., due to mobility) since initial assumptions of the environment are violated or stated too general. In this paper, we propose ADAPTIVECEP, a CEP system that is able to self-adapt to detected changes in environmental conditions. We propose a CEP query language that allows specifying changes in the behavior of the CEP system and its mechanisms in detecting events dependent on environmental conditions. This way, ADAPTIVECEP can select the best-suited configurations for given quality demands. In our evaluation, we show by means of a reference concept how the flexibility exposed by the query language helps to achieve significant performance gains.

Language-integrated privacy-aware distributed queries

“…To reduce the load of the placement algorithm, Zhou et al propose a coarser granularity (Zhou et al 2006) and deploy query fragments, i.e., groups of operators. AdaptiveCEP (Weisenburger et al 2017) enables developers to define their own placement strategies. In contrast to theses solutions and others (Tian and DeWitt 2003;Xing et al 2005) we adopt static operator placement because we derive placement from privacy constraints which are statically defined.…”

INetCEP: In-Network Complex Event Processing for Information-Centric Networking

“…Data Plane and Query Languages. Many CEP systems propose a query language that supports users to specify and embed the features provided in their applications [50]. A particular, challenge when supporting network architectures is to be able to decompose the query specification such that it can be mapped to the programming models provided by dedicated network architectures.…”

GRASS: generic reactive application-specific scheduling

“…Similar to our approach of specifying resource scheduling policies by leveraging reactive programming, FRP has been applied to compute the operator placement for complex event processing operators [36]. Recent research on RP focuses on issues such as concurrency [11], fault tolerance [25], distribution [35,37], different levels of consistency [23] and the application to areas such as autonomous vehicles [13] and IoT and edge computing [5].…”