 |
|
 |
Verteilte Systeme: Forschung
|
|
Projekt DHEP
|
|
|
|
|
|
| |
|
Distributed Heterogeneous Event Processing
|
| Projekttyp |
CAS Projekt |
| Gefördert durch |
IBM Deutschland Research & Development GmbH |
| Beginn |
03.01.2008
|
| Ende |
28.02.2010
|
| Leiter |
Prof. Dr. rer. nat. Dr. h. c. Kurt Rothermel
|
| Mitarbeiter |
Dr. Boris Koldehofe
Dipl.-Inform. Björn Schilling
|
| Ansprechpartner |
Dr. Boris Koldehofe
Dipl.-Inform. Björn Schilling
|
| Kooperationspartner |
IBM Deutschland Research & Development GmbH
|
| Kurzbeschreibung |
Context and objective
“Distributed Heterogeneous Event Processing” is a project under the Center of Advanced Studies program for which IBM funds research activities at various universities. The project is also aligned with the IBM Technology Partnership Center initiative between the Information Science faculty of the Universität Stuttgart and IBM Deutschland Research & Development GmbH. It involves researchers from the IBM lab and the IPVS (Institute of Parallel and Distributed Systems) to investigate high performance and scalable complex event processing in the context of business and manufacturing solutions.
Complex Event Processing
Complex event processing is a technology which allows correlating basic events to complex events closely aligned with the semantics of business processes related to the events. Climbing up the ladder from technical to business events, complex event processing decouples providers of technical information providers and consumers of semantically rich information. As a consequence, changing technical event sources will not affect the delivery of semantically rich events to business applications. By performing correlations close to the data sources, complex event processing also contributes in increasing the efficiency of event processing by reducing the amount of data to be propagated to the upper layers of the network. Complex Event Processing is an emerging paradigm of enterprise computing and IBM considers it as the key technology in many applications and business solutions. Moreover, IBM is one of the key players in providing advanced correlation technology implementations through respective event correlation engines.
Project Scope
Most solutions in industry implement complex event processing by performing correlations in a centralized fashion. However, many applications such as manufacturing systems or Radio Frequency Identification solutions are highly demanding in terms of the generated event load itself, as well as in terms of the computational complexity which requires more hardware capacity than just a single event correlation server. Furthermore, if correlation of events happens over multiple domains, centralized solutions are expected to perform badly with respect to response times and are vulnerable to availability of event-driven information processing.
The project ‘Distributed Heterogeneous Event Processing’ explores new ways in order to increase the scalability and availability of existing complex event processing technology by establishing a distributed event correlation network. In particular, the focus is on the s pecification of event processing rules for business domains like manufacturing or retail and the development of corresponding deployment strategies that allow for dynamic placement of correlation functionality on multiple event correlation nodes. Depending on the application requirements, different correlation engines can be used across an event correlation network. To support the efficient deployment and management of correlation networks, the project investigates approaches to optimize and automate the placement of interacting event correlation rules across network nodes.
|
| |
|
|
|
|
|
Björn Schilling
)
| © Universität Stuttgart |
Impressum