Event processing will play an increasingly important role in constructing enterprise applications that can immediately react to business critical events. Various technologies have been proposed in recent years, such as event processing, data streams and asynchronous messaging (e.g. pub/sub). We believe these technologies share a common processing model and differ only in target workload, including query language features and consistency requirements. We argue that integrating these technologies is the next step in a natural progression. In this paper, we present an overview and discuss the foundations of CEDR, an event streaming system that embraces a temporal stream model to unify and further enrich query language features, handle imperfections in event delivery, define correctness guarantees, and define operator semantics. We describe specific contributions made so far and outline next steps in developing the CEDR system.

Evaluation of complex event queries over time involves storing information about those events that are relevant for, i.e., might contribute to, future answers. We call the period of time for which an event or an intermediate result must (at least) be stored its temporal relevance. This paper pioneers a precise definition of temporal relevance and develops a method for statically (i.e., at compile time) determining it. During query evaluation (i.e., at run time), this enables garbage collection of events that become irrelevant as time progresses. Temporal relevance is also important at compile time for cost-based query planning.

In this report we use the kell-m process algebra to develop three models for Distributed Event-Based Systems (DEBSs). The first model is of the DEBS API standard proposed by Pietzuch et al. The second model is for the hierarchical structuring mechanism for components in the REBECA DEBS. The third model is for the internal structure of administrative components in the NaradaBrokering DEBS. These models support the specification of DEBS properties previously proposed in the area using other formalisms. We also show how new properties, based on the locality features provided by kell-m and the ability to passivate kells, can now be specified. 1

Event detection is an important aspect of many application types, ranging from active databases over digital libraries and stock market agents, to reactive embedded systems. To allow systems to react to complex events patterns rather than to simple primitive events, an event algebra can be used. We have developed an event algebra with formal semantics and a number of useful algebraic properties. An equivalent operational semantics has been developed, and we have identified an important subset of event expressions for which the detection mechanism can be implemented with constant memory.

Abstract not found

Systems for distributed event processing have recently gained increasing attention in a broad range of application domains. This raises the demand for methods to adapt the system design to application‐ specific needs. Our approach considers (i) tradeoffs regarding the hardware infrastructure and (ii) tradeoffs in the software design. The presented model supports design decisions in dependence of application‐specific event properties and design goals.

Abstract not found

Event‐based systems (EBS) have received increasing attention in the past decade from various communities. Central to these systems is the notion of event, which is often considered as ”a happening of interest”. An EBS encompasses a large range of functionalities on various technological levels (e.g., language, execution, or communication). Many approaches exist to handle events and the solutions are not always either defined on a clear level or well understood. This is mainly due to the difference between the communities that tackle the issues, which also leads to some confusion in the terminology. The notion of events itself is often not clearly defined and varies from one approach to the other. This paper presents Mātārere, a layer decomposition of a generic EBS into levels of abstraction. The goal of the solution proposed here is threefold: (1) provide unifying terminology, (2) understand better the various existing approaches which take place on different levels, and (3) offer some support to build an EBS using this uniform framework. A strength of our approach lies in the consideration of recursion ‐ in the sense that all layers may repeat themselves on

Remove the contents below from this page before submission

Data stream management systems (DSMS) so far focus on event queries and hardly consider combined queries to both data from event streams and from a database. However, applications like emergency management require combined data stream and database queries. Further requirements are the simultaneous use of multiple timestamps after different time lines and semantics, expressive temporal relations between multiple time-stamps and flexible negation, grouping and aggregation which can be controlled, i. e. started and stopped, by events and are not limited to fixed-size time windows. Current DSMS hardly address these requirements. This article proposes Temporal Stream Algebra (TSA) so as to meet the afore mentioned requirements. Temporal streams are a common abstraction of data streams and database relations; the operators of TSA are generalizations of the usual operators of Relational Algebra. A in-depth ’analysis of temporal relations guarantees that valid TSA expressions are non-blocking, i. e. can be evaluated incrementally. In this respect TSA differs significantly from previous algebraic approaches which use specialized operators to prevent blocking expressions on a “syntactical ” level.