Supporting the dynamic grid service lifecycle

Abstract

This paper presents an architecture and implementation for a dynamic OGSA-based Grid service architecturethat extends GT3 to support dynamic service hostingwhere to host and re-host a service within the Grid in response to service demand and resource fluctuation. Our model goes beyond current OGSI implementations in which the service is presumed to be pre-installed at all sites (and only service instantiation is dynamic). In dynamic virtual organizations (VOs)% we believe dynamic service hosting provides an important flexibility. Our model also defines several new adaptive Grid service classes that support adaptation at multiple levels. Dynamic service deployment allows new services to be added or replaced without taking down a site for reconfiguration and allows a VO to respond effectively to dynamic resource availability and demand. The preliminary results suggest that the cost of dynamic installation, deployment, and invocation, is tolerable'. 0-7803-9074-

Two Ways to Grid: The Contribution of Open Grid Services Architecture (OGSA) Mechanisms to Service-Centric and Resource-Centric Lifecycles

“…However, integration of these mechanisms with a cross-organizational deployment capability (e.g. to reactivate passivated services on nodes across firewalls) would be a major hurdle, although some research has started to address this [16,17,18].…”

HAND: Highly Available Dynamic Deployment Infrastructure for Globus Toolkit 4

“…Weissman et al [2] have implemented dynamic deployment in GT by building on the dynamic deployment-enabled Apache Tomcat server as the hosting environment. With this approach, users package their services with the basic Java WS Core libraries as a WAR file and deploy it into Tomcat as a Web application.…”

“…Experience within ChinaGrid [1,9] and elsewhere [2,17] has emphasized the importance of dynamic service deployment and management as an enabler of dynamically extensible virtual organizations (VOs) [8]. More specifically, we identify the following requirements when services are hosted on a dynamic management-enabled Grid:…”

“…Dynamic deployment is a big challenge: frequent shutting down and starting up of services for software upgrades or changes in resources or users can increase management costs. Some dynamic deployment solutions have been proposed based on Apache Tomcat's dynamic deployment functionality [2,3,14]. However, that infrastructure provides poor performance and availability, with the consequence that (for example) dynamic deployment during a workflow's execution can cause a critical task to fail.…”

“…And in both cases, the polynomial time are all O(x 2 ). To enable direct comparison with other dynamic deployment enabled containers, we repeated our experiments on the Apache Tomcat container (version 5.0.30), as used by Weissman et al [2,3] and Matthew et al [17]. (However, we note that we use GT4, not GT3 as used by those authors.)…”