No abstract

No abstract

Meeting client Quality-of-Service (QoS) expectations proves to be a difficult task for the providers of e-Commerce services, especially when web servers experience overload conditions, which cause increased response times and request rejections, leading to user frustration, lowered usage of the service and reduced revenues. In this paper, we propose a server-side request scheduling mechanism that addresses these problems. Our Reward-Driven Request Prioritization (RDRP) algorithm gives higher execution priority to client web sessions that are likely to bring more service profit (or any other application-specific reward). The method works by predicting future session structure by comparing its requests seen so far with aggregated information about recent client behavior, and using these predictions to preferentially allocate web server resources. Our experiments using the TPC-W benchmark application with an implementation of the RDRP techniques in the JBoss web application server show that RDRP can significantly boost profit attained by the service, while providing better QoS to clients that bring more profit.

Recent studies showed potential for using component frameworks for building flexible adaptible applications for deployment in distributed environments. However this approach is hindered by the complexity of deployment of component-based applications, which usually involves a great deal of configuration of both the application components and system services they depend on. In this paper we propose an infrastructure for automatic dynamic deployment of J2EE applications, that specifically addresses the problems of (1) inter-component connectivity specification and its effects on component configuration and deployment; and (2) application component dependencies on application server services, their configuration and deployment. The proposed infrastructure provides simple yet expressive abstractions for potential application adaptation through dynamic deployment and undeployment of components. We implement the infrastructure as a part of the JBoss J2EE application server and test it on several sample J2EE applications.

Distributed deployment of network applications in widearea environments has proven effective for improving enduser experience. Another trend is the use of component frameworks for building network services. Their component-based nature makes such applications natural candidates for distributed deployment, but it is unclear if the design patterns underlying component frameworks also enable efficient service distribution.In this paper, we investigate the application design rules and accompanying system-level support essential to a beneficial and efficient service distribution process. Our study targets the widely used Java 2 Enterprise Edition (J2EE) component platform and Java Pet Store, a sample component-based e-commerce application.Our results present strong experimental evidence that component-based applications can be efficiently distributed in wide-area environments using a small set of generallyapplicable design rules for orchestrating interactions and managing component state. We additionally discuss enforcement of these rules, and their automated implementation by container frameworks.

Meeting client QoS expectations proves to be a difficult task for the providers of modern Internet services, especially when web servers experience overload conditions (steady or transient), which cause increased response times and request rejections, leading to user frustration, lowered usage of the service and reduced revenues. In this paper, we propose a server-side request scheduling mechanism that addresses these problems. Our Reward-Driven Request Prioritization (RDRP) algorithm gives higher execution priority to client web sessions that are likely to bring more service profit (or any other application-specific reward). The method works by predicting future session structure by comparing its requests seen so far with aggregated information about recent client behavior, and using these predictions to preferentially allocate web server resources. Our experiments using the TPC-W benchmark application and CBMG-based web workloads, with an implementation of the RDRP techniques in the JBoss application server, show that RDRP can significantly boost reward attained by the service, while providing better QoS to clients that bring more reward.

Deployment patterns have been proposed as a mechanism to support the provisioning of SOA-based services. Deployment patterns represent the structure and constraints of composite solutions, including non-functional properties, such as performance, availability, and security, without binding to specific resource instances. In previous work [1], we have presented a formal mechanism for capturing such service deployment patterns using models. Our pattern models define abstract connectivity and configuration requirements which are then realized by an existing or planned infrastructure. Realization mapping is used to enforce policies, and is materialized at deployment time. In this paper we extend that work to address the problem of automatic pattern realization over a given infrastructure. We first formalize the problem and present three variations of increasing power and complexity. We then present a variation of a search-based graph isomorphism algorithm with extensions for our pattern model semantics. Next, we show that our worst-case exponential complexity algorithm performs well in practice, over a number of pattern and infrastructure combinations. We speculate that this is because deployment topologies represent heavily labeled and sparse graphs. We present a number of heuristics which we have compared experimentally, and have identified one which performs best across most scenarios. Our algorithm has been incorporated into a large deployment modeling platform, now part of the IBM Rational Software Architect (RSA) tool [2].

No abstract