Advances in Quality of Service Management, EDOC 2008

AQuSerM: Advances in Quality of Service Management

EDOC 2008 workshop, 18th September, 2008, München, Germany

This workshop's theme is advances in QoS-oriented techniques and tools for managing enterprise architectures, encompassing approaches to monitoring, diagnostics, runtime analysis and prediction and adaptation. Serviced-oriented and model-driven approaches will be a special focus of the workshop.

AQuSerm 2008 continues a successful and stimulating series of workshops, held in 2007 and 2006. An associated special issue of the Journal of Object technology is available here.

Workshop goals

Service Level Management (SLM) is the process of managing the Quality of Service (QoS) demanded by clients and offered by providers. In the past, SLM approaches have focused on service contract definition, monitoring and reporting and have typically been handled by enterprise system management tools such as Microsoft’s SMS, CA’s Unicenter and Empirix’s OneSight.

However, traditional approaches are inadequate when dealing with complex service-oriented architectures. Service-oriented architectures are compositional, dynamic and often distributed over the internet. For such architectures, SLM becomes a difficult problem that can no longer be handled by traditional monitoring tools. This is because of the dynamic, flexible, compositional and global natures of SOAs.

This workshop will be concerned with the issues that are important to modern QoS management: the monitoring of widely distributed components, dynamic adaptation strategies and the necessity for more sophisticated prediction and diagnostic analysis techniques. Model-driven approaches to these issues will be a special focus of the workshop.

The workshop shall bring together researchers from academia and industry interested in cutting edge formal and model-based approaches as well as utilizing current standards and middleware to meet the challenges of SLM for the 21st century.

Themes

The main theme of the workshop will be QoS-oriented techniques and tools for managing modern enterprise architectures, encompassing approaches to monitoring, diagnostics, runtime analysis, behaviour prediction, adaptation strategies and the interrelation of these issues within SLM.

Special focus will be given to model-driven approaches. The development of standards such as the ISO/IEC QoS Framework, the RM-ODP and the UML Profile for QoS are intended to form the basis for the design and implementation of QoS management in networked enterprise architectures. A current open question is how best to use these standards within the Model Driven Architecture (MDA) refinement strategy for software development. For example, some authors are advocating the use of MDA to generate platform-specific monitorable implementations from QoS requirements specified in a platform-independent metamodel.

Topics include, but are not limited to,

QoS support for common enterprise middleware (such as CORBA, J2EE and .NET)
Management issues for domain-specific architectures (such as process control architectures built using OPC)
Managing complex systems using current industrial monitoring infrastructures and standards (e.g., Microsoft’s WMI and the DMTF’s CIM standard)
Service-oriented approaches to QoS management
Formal methods to support SLM
Mathematical models for system diagnostics
Industrial SLM case studies
Model-driven approaches to monitoring, diagnostics, prediction and adaptation
Unifying management frameworks

Preliminary Programme

	Thursday 18 September 2008, LMU
8:30-9:00	Registration
9:00-10:00	Invited Talk (main EDOC event) Pradeep Ray
10:00-10:30	Break
10:30-10:45	Opening remarks Iman Poernomo and Guijun Wang, Quality of Service Managment
10:45-12:15	Session 1:Models for QoS Management Sumant Tambe, Akshay Dabholkar, Aniruddha Gokhale and Amogh Kavimandan (Vanderbilt University): Towards A QoS Modeling and Modularization Framework for Component-based Systems Paper. Current domain-speciﬁc modeling (DSM) frameworks for designing component-based systems provide modeling support for system’s structural as well as non-functional or quality of service (QoS) concerns. However, the focus of such frameworks on system’s non-functional concerns is an after-thought and their support is at best adhoc. Further, such frameworks lack strong decoupling between the modeling of the system’s structural composition and their QoS requirements. This lack of QoS modularization limits (1) reusability of such frameworks, (2) ease of their maintenance when new non-functional characteristics are added, and (3) independent evolution of the modeling frameworks along both the structural and non-functional dimensions. This paper describes Component QoS Modeling Language (CQML), which is a reusable, extensible, and platform-independent QoS modeling language that provides strong separation between the structural and non-functional dimensions. CQML supports independent evolution of structural metamodel of composition modeling languages as well as QoS metamodel. To evaluate, we superimpose CQML on a purely structural modeling language and automatically generate, conﬁgure, and deploy component-based fault-monitoring infrastructure using aspect-oriented modeling (AOM) techniques. Javier F. Briones, Miguel Ángel de Miguel, Alejandro Alonso, Juan Pedro Silva (Universidad Politécnica de Madrid): Modeling Quality of Service Adaptability Paper. Quality of service adaptability refers to the ability of services (or components) to adapt the quality exhibited during run-time, or to the faculty of architectural models to show that several alternatives concerning quality could be implemented. Enclosing quality properties with architectural models has been typically used to improve system understanding. Nevertheless, these properties can also be used to compose subsystems whose quality can be adapted or/and to predict the behavior of the run-time adaptability. Existing software modeling languages lack enough mechanisms to cope with adaptability, e.g. to describe software elements that may offer/require several quality levels. This paper presents concepts that such a language needs to include to model quality adaptable systems, and how we use those concepts to compose and analyze software architectures. Session 2: Management and Instrumentation Christof Momm, Thomas Detsch and Sebastian Abeck (FZI, Universität Karlsruhe): Model-Driven Instrumentation for Monitoring the Quality of Web Service Compositions Paper. Supporting business services through Web service compositions (WSC) as part of service-oriented architectures involves various runtime monitoring requirements. The implementation of these requirements results in additional development activities. Due to the lack of standards for treating such WSC monitoring concerns, a corresponding development approach has to deal with a variety of specific technologies. This paper therefore introduces a platform-independent approach to the instrumentation of WSC and the generation of an effective monitoring infrastructure based on the principles of model-driven software development (MDSD).
12:15-13:45	Lunch
13:45 - 15:45	Session 2 (continued): Management and Instrumentation Rodolfo Santiago, Guijun Wang, Harley Chen and Changzhou Wang (Boeing): Interoperability of End to End Quality of Service (QoS) Management across Heterogeneous Platforms in System of Systems Paper. Interconnectivity, interoperability, evolution and emergent behavior are some of the key characteristics of System of Systems (SOS). Providing end to end Quality of Service (QoS) for an SOS is a major challenge today because of issues with connectivity, the independent evolution of component systems and their heterogeneity. Heterogeneity of systems is common because of the independent evolution paths and development cycles of the systems. In this paper, we describe an interoperable end to end QoS management solution including concepts, architecture, prototype, and experimental results. Our solution not only provides QoS at systems level, but also ensuring end-to-end QoS across heterogeneous systems at the SOS level. To validate our solution, we analyzed and compared architecture alternatives and developed prototype and experiments employing multiple and varied data dissemination technologies. Session 3: Automated Control Shinji Kikuchi, Yohsuke Isozaki, Yoshihiro Kanna (NEC Corporation): A Proposal of the New Architectural and Mathematical Control Model for Maintaining SLA of Composite Web Services Paper. Recently, the certain demand in regards to available elemental services provided by other independent firms for compositing new services has increased. However, under current status, whenever it is hard to maintain the required level of quality of the new composite web service, assignment of the new computer’s resources is not always effective especially for the composite web service providers. Thus, a new approach might be required. This paper presents a proposal of a new control method for the composite web services. Here, following two items are applied; firstly, the theory of constraints proposed by Israeli physicist E.M.Goldratt, secondly the evaluation process in the feed forward controlling method. Here, the architectural and theoretical aspects of the method are defined. Assel Akzhalova, Mahbub Gani and Iman Poernomo (King's College London): Runtime adaptation of queued networks through dynamic programming Paper. Self-adaptive systems are capable of changing their behaviour at runtime to meet target behavioural constraints. An important research question is how quality of service models can inform runtime adaptation. This talk presents a step towards solving this problem by application of dynamic programming. By utilizing dynamic programming, our new approach has the advantage of determining an optimal control solution directly from a given discrete set of adaptation policies. Constandinos X. Mavromoustakis (University of Nicosia) and Helen D. Karatza (Aristotle University of Thessaloniki) End-to-end layered asynchronous scheduling scheme for energy aware QoS provision in asymmetrical wireless devices Paper. In order to establish and maintain a Quality of Service (QoS) framework in wireless network, a major requirement is to prolong network lifetime, with simultaneous preservation of other major requirements like coverage, connectivity and adequate resources. In this work an approach is proposed which on a periodic basis it schedules the interfaces into active and inactive phases by using a layered state model and a tiered-based architecture. The model takes into account metrics like the signal strength and devices’ capacity limitations, and encompasses them into the layered asynchronous scheduling scheme for end-to-end reliability and performance measures extraction. Each device schedules the next state (active, semi-sleep or sleep state) according to traffic and channel data rate measures, combined with the layered state scheme. Through the designed tiered architecture, and through experimental simulation, the proposed QoS energy-aware management scheme is thoroughly evaluated in order to meet the parameters’ values where the optimal QoS and throughput response for each device/user is achieved.
15:45-16:00	Concluding remarks and discussion

Important dates

Paper submission deadline: 20 June 2008
Paper acceptance notification: 21 July 2008
Camera ready of papers: 22 August 2008
Workshop date: 18 September 2008

Submissions

Submissions should be 4 to 10 pages long in IEEE Computer Society format and include the author's name, affiliation and contact details. They should be submitted by e-mail as PDF files to the Workshop Chairs (iman.poernomo ' at symbol ' kcl.ac.uk or guijun.wang ' at symbol ' pss.Boeing.com) by the deadline.

At least one author of an accepted paper should participate in the workshop.

All accepted papers for the workshop will appear in the IEEE Digital Library. Authors will be invited to extend their papers for a special issue of a leading international journal (this is currently under negotiation). Revised best papers of previous AQuSerMs were published in the Journal of Object Technology.