Christian Schröpfer

Negotiating and Enforcing QoS and SLAs in Grid and Cloud Computing

Emerging grid computing infrastructures such as cloud computing can only become viable alternatives for the enterprise if they can provide stable service levels for business processes and SLA-based costing. In this paper we describe and apply a three-step approach to map SLA and QoS requirements of business processes to such infrastructures. We start with formalization of service capabilities and business process requirements. We compare them and, if we detect a performance or reliability gap, we dynamically improve performance of individual services deployed in grid and cloud computing environments. Here we employ translucent replication of services. An experimental evaluation in Amazon EC2 verified our approach.

Introducing Preferences over NFPs into Service Selection in SOA

When implementing a business or software activity in SOA, a match is sought between the required functionality and that provided by a web service. In selecting services to perform a certain business functionality, often only hard constraints are considered. However, client requirements over QoS or other NFP types are often soft and allow tradeoffs. We use a graphical language for specifying hard constraints, preferences and tradeoffs over NFPs as well as service level objectives (SLO). In particular, we use the TCP and UCP network formalisms to allow for a simple yet very flexible specification of hard constraints, preferences, and tradeoffs over these properties. Algorithms for selecting web services according to the hard constraints, as well as for optimizing the selected web service configuration, according to the specification, were developed.

Service-level enforcement in web-services-based systems

Web-services-based systems present challenges for service-level enforcement. Contrary to typical enterprise architectures where we are in control of all services and can easily specify and monitor service levels, here we need self-adapting approaches. This article proposes an approach for the assurance of service levels. The steps of the approach are: the formalisation of requirements using a Service Level Objective (SLO), the formalisation of web service properties in a service-level capability statement and adaptive actions based on the translucent replication of services. The evaluation is based on a reference process model in Business Processing Modelling Notation (BPMN) and subsequent mapping at the infrastructure level. This mapping includes the definition, implementation and deployment of web services that cover the functional requirements, orchestration and coordination of these services using Business Process Execution Language (BPEL), as well as the the runtime monitoring and assurance of Nonfunctional Properties (NFPs) within a business process management framework.

Techniques for service level enforcement in web-services based systems

Web-services based systems present challenges for service level enforcement. Contrary to typical enterprise architectures where we are in control of all services and can easily specify and monitor service levels, here we need self-adapting approaches. The paper proposes techniques for service level enforcement in the context of a four step approach. The proposed techniques are: formalization of requirements using a service level objective, formalization of Web Service properties in a service level capability statement, and adaptive actions based on translucent replication of services. For an experimental evaluation we deployed a Web Service benchmark and our replication framework in Amazons Elastic Compute Cloud. It provided continuous meeting of specified service levels.

A Flexible Approach to Service Management-Related Service Description in SOAs

In order for service-oriented architectures (SOAs) to deliver their true value for the business, e.g. flexibility and transparency, a holistic service management needs to be set up in the enterprise. To perform all the service management tasks efficiently heavy support by automated processes and tools is necessary. This article describes a service description approach that is based on OWL-S (Web Ontology Language for Services) and focuses on nonfunctional criteria. It starts with the necessary service management tasks and explains non-functional data elements and statements for its automated support. After covering related work it explains the proposed flexible extension to OWL-S. This extension is twofold. Firstly, simple service lifecycle elements are added using the normal extension mechanism. Secondly for adding QoS (Quality of Service) capabilities, the approach combines this extension mechanism with UML (Unified Modeling Language) Profile for QoS. A prototype delivers the proof-of-concept.

Implementing non-functional service descriptions in SOAs

This article describes a framework for extended service descriptions based on OWL-S (Web Ontology Language for Services) focusing on nonfunctional criteria. Necessary service management tasks will be introduced and extended by corresponding data elements and statements for its automated support. After a short comparative description of several existing approaches to semantic service descriptions the paper addresses the actual extension of OWL-S. Non-functional extensions as service lifecycle elements and Quality of Services (QoS) are added. To extend QoS capabilities, the approach combines the common extension mechanism with UML (Unified Modeling Language) Profile for QoS. A prototype delivers the proof-of-concept for the first part of the extension. The prototype implements SOA-specific authentications and all basic features for a tool-supported service management using extended semantic service descriptions by defining an ontology-based service taxonomy and service annotation.

Introducing a Method to Derive an Enterprise-Specific SOA Operating Model

Considering that each IT-organization is based on an established IT-governance the introduction of SOA as a new overall architectural approach will change the requirements for the specific governance model. The paper presents a SOA-specific governance model which differentiates an operating model, service lifecycle management, service ownership, cost allocation, and meta-data management support. Based on the companys maturity along the SOA-relevant dimensions IT-governance, SOA experience, and process maturity, its size, as well as its organizational structure the paper discusses an SOA operating model with organizational structure, roles, responsibilities, and KPIs. The main assumption of the approach is the alignment of the organizational structure of IT with a functional domain model. This operating model is the outcome of ten industrial workshops and one complex case study based on standards as CobiT and ITIL.

Extending the UN/CEFACT modeling methodology and core components for intra-organizational service orchestration

When creating a companys IT structure based on a service-oriented architecture (SOA), it is necessary to first analyze the business domains and process areas of the company, then to model the business processes to be supported by the SOA and finally to convert the models into a service orchestration description. Currently, few methodologies exist to support this. At our department, we have proven that the UN/CEFACT Modeling Methodology (UMM) can be used for intra-organizational process integration. In this article we analyze if the UMM is sufficient for SOA, which artifacts are missing and how the UMM could be extended. The UMM was created to model the collaboration between different legal entities to perform collaborative business processes. There exist methods to convert these models into executable service choreography descriptions expressed in the Business Process Specification Schema (BPSS) or the Business Process Execution Language (BPEL). However, the business process models can also be used as a basis for an intra-organizational service orchestration. By extending the UMM it is possible to enable the automated generation of service orchestrations using Core Components and the Business Process Modeling Notation (BPMN).