QoS in Grid Computing

QoS in Grid Computing

Zhihui Du (Tsinghua University, China), Zhili Cheng (Tsinghua University, China), Xiaoying Wang (Tsinghua University, China) and Chuang Lin (Tsinghua University, China)
DOI: 10.4018/978-1-60566-184-1.ch008
OnDemand PDF Download:
No Current Special Offers


This chapter first summarizes popular terms of QoS related concepts and technologies in grid computing, including SLA, End-to-End QoS Provision and Virtualization. Then a three layered general grid QoS provision model based on MetaServices is proposed. Operating mechanisms are discussed in detail, and the model can maintain grid QoS by defining QoS requirements in different levels and solve the QoS problems hierarchically. A prototype named PMGrid is designed and implemented based on the QoS provision model. PMGrid is a grid system for astronomy data processing. The results show that the PMGrid can maintain the QoS requirements of astronomy data processing.
Chapter Preview


Before the descriptions of the Grid QoS provision model later in the part of “main focus”, technology hotspots of Grid QoS provision are discussed as the background in this section, including: Grid QoS negotiation protocols designed based on SLA, such as SNAP; End-to-End QoS Provision, concept borrowed from networks (Sander et al 2001); and Virtualization technologies which have been introduced into the area of Grid computing recently.

Agreement Protocols

In order to maintain a QoS-guaranteed communication path between services, the Service Level Agreement (SLA) is employed into Grid environment.

An SLA is a formal negotiation agreement between two parties. It is a contract that exists between customers and service providers. It records the QoS requirement of services, priorities, responsibilities, guarantee etc. (Wikipedia – Service Level Agreement).

In Grid computing, SLAs are instantiated via the Service negotiation and Acquisition Protocol (SNAP), which provides contract lifetime management (Czajkowski et al., 2002). SNAP proposes three different types of SLAs: Task service level agreements (TSLA), Resource service level agreements (RSLA), and Binding service level agreements (BSLA). It also defines an internal bundling model. Based on the standard form of SNAP, various Grid services keep consulting and interacting. For more detailed information of SNAP, please refer to the relevant descriptions in (Czajkowski et al 2002).

Key Terms in this Chapter

MetaService: Services abstracted to identify the essential management functions which should be provided in an OGSA based Grid environment.

Grid QoS: A set of quality requirements on the collective behavior of one or more services in Grid environment to prioritize different requirements.

Resource Reservation Protocol (RSVP): A network layer protocol designed to reserve resources across a network for integrated services over Internet.

Service Level Agreement: An agreement describes the user requirements for QoS by XML-based description language. It is defined in a unified service interface in order to achieve QoS service management, consultation and agreement.

Virtual Workspace: A framework, which allows a Grid client to define an environment in terms of requirements (such as resource requirements or software configuration), manage it, and deploy the environment in the Grid environment (Keahey and Galron et al., 2005).

Grid Middleware: Software designed to standardize and to ensure that the implementation of Grid application complied with Grid unified norms.

Virtual Machine: A number of discrete identical execution environments on a single computer, each of which runs an operating system (OS). This can allow applications written for one OS to be executed on a machine which runs a different OS, or provide execution “sandboxes” which provide a greater level of isolation between processes than is achieved when running multiple processes on the same instance of an OS.

Complete Chapter List

Search this Book: