Distributed OSGi built over message‐oriented middleware

Psiuk, Marek; Żmuda, Daniel; Zieliński, Krzysztof

doi:10.1002/spe.1148

Cited by 5 publications

(1 citation statement)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The testbed on which the approach was evaluated consisted of three OSGi containers: one with the middleware services; the application services were distributed over two remaining OSGi containers. The communication between separated OSGi containers was implemented using the mechanisms presented in our earlier work [14].…”

Section: Discussionmentioning

confidence: 99%

Middleware for Managing QoS Adaptation of SOA Applications

Psiuk

Zieliński

2015

CMST

Self Cite

View full text Add to dashboard Cite

The paper describes an improvement over our previous work: the concept of an Adaptive SOA Solution Stack (based on the IBM S3 model) which applies an AS3 element pattern to S3 layers where the need for adaptation arises. The presented improvement, called the adaptation strategy management process, represents a solution that enables an Adaptation Architect to model Quality of Service (QoS) adaptation in a declarative manner, automatically deploy it into a running system and then monitor its execution. Its main objective is to allow the Adaptation Architects to view the adaptation process on a higher level of abstraction and employ adaptivity mechanisms in working applications in an easy way. This is accomplished by incorporating the adaptive application metamodel developed in the DiVA EU project and adjusting it to the SOA context. This paper explains the challenges involved in adaptation strategy management and proposes extensions to the DiVA metamodel. Subsequently, it presents a method by which the Adaptive Manager (a component of the AS3 adaptation loop responsible for making decisions about adaptation) can execute adaptation strategies in accordance with the adaptation model. The presented approach is evaluated in a case study, creating an adaptation strategy and monitoring its impact on an application prototype.

show abstract

Section: Discussionmentioning

confidence: 99%