Web Services Dependability

Web Services Dependability

Michael C. Jaeger (Berlin University of Technology, Germany) and Matthias Werner (Operating Systems Group, TU Chemnitz, Germany)
Copyright: © 2009 |Pages: 17
DOI: 10.4018/978-1-60566-042-4.ch007
OnDemand PDF Download:


This chapter presents the definition of relevant terminology and a conceptual model of the basic terms. The chapter starts with the presentation of research in the area of dependability. Based on this, Web service concepts related to the dependability are introduced. The presentation leads into a statement identifying individual quality-of-service (QoS) characteristics for forming dependable Web services. Then, the chapter discusses the current status quo in the area of QoS-aware Web services. This part is divided into three subparts: description, management, and monitoring. This also identifies ongoing efforts as well as efforts that do not show present activity. By this discussion, this chapter puts research about dependability in relation with ongoing QoS-related efforts in the Web services domain. Thus it identifies how these efforts can be combined in order to form a dependable Web services platform.
Chapter Preview


The Web services as an implementation of the service oriented architecture (SOA) represents a trend in the IT industry for the development of a flexible and unifying software infrastructure. Picking up the Web services idea, software components provide their functionality as a service by using uniform interface description and invocation protocols using XML formats and Internet communication protocols.

One important characteristic in an open, distributed system is that services can appear, disappear, and bound dynamically at run time. The deployment and use of Web services also results in a varying level of quality. The access to Web services can involve the Internet and thus relies, as mentioned above, on its protocols. As the Web pages in the Internet, Web services are provided on a best-effort basis as well. Indeed, Internet protocols are known for their robustness. However, they do not guarantee that a connection can be established and can be held at given characteristics. Thus, for some soft real-time applications such as telephony or brokering stocks, reservation and signalling extensions to the Internet protocols are used, making Web services more reliable. In addition, the basic SOA idea of a loosely coupled provision of services, which also applies to Web services, forms also a problem. The binding of services leaves generally open what time it takes to find a service, to bind a service, and how reliable the service execution will perform. Thus, the use of loose-coupling mechanisms poses a problem regarding a satisfactory overall performance.

These quality of service (QoS) issues with Web services are well known. Thus, several recent developments and approaches have been proposed that extend the Web services with individual attributes, such as security, integrity, reliability, availability, and further characteristics. Today, a large number of proposals and efforts exist that provide research on description formats for QoS-aware Web services (e.g., Tosic, Patel, & Pagurek, 2002; Ludwig, Keller, Dan, King, & Franck, 2003; Ran 2003; Zhou, Chia, & Lee, 2004), a management infrastructure (e.g., Sahai et al., 2002; Wang, Chen, Wang, Fung, & Uczekaj, 2004), and monitoring approaches (e.g., Baresi, Guinea, & Plebani, 2005; Sahai et al., 2002; Tosic, Ma, Pagurek, & Esfandiari, 2004). In the recent years, standards and recommendations from nonprofit organisations have also emerged to form a community-wide consensus for establishing reliable Web services (e.g., WS-reliability) (OASIS, 2004). In summary, the different QoS issues have already been extensively covered in the Web services domain.

Complete Chapter List

Search this Book: