Many e-Science applications are increasingly relying on orchestrating workflows of static web ser... more Many e-Science applications are increasingly relying on orchestrating workflows of static web services. The static nature of these web services means that workflow management systems have no control over the underlying mechanics of such services. This lack of control manifests itself as a problem when optimizing workflow execution since techniques such as data-locality aware deployment and service-to-service communication are very difficult to achieve. In this paper we propose a novel approach for mobilizing scientific web services onto common distributed resources and as such enable back-to-back communication between cooperating web services, autonomous web service scaling through fuzzy control and autonomous web service workflow orchestration.
Document coherency is a challenging problem for Web caching. Once the documents are cached throug... more Document coherency is a challenging problem for Web caching. Once the documents are cached throughout the Internet, it is often difficult to keep them coherent with the origin document without generating a new traffic that could increase the traffic on the international backbone and overload the popular servers. Several solutions have been proposed to solve this problem, among them two categories have been widely discussed: the strong document coherency and the weak document coherency. The cost and the efficiency of the two categories are still a controversial issue, while in some studies the strong coherency is far too expensive to be used in the Web context, in other studies it could be maintained at a low cost. The accuracy of these analysis is depending very much on how the document updating process is approximated. In this study, we compare some of the coherence methods proposed for Web caching. Among other points, we study the side effects of these methods on the Internet traffic. The ultimate goal is to study the cache behavior under several conditions, which will cover some of the factors that play an important role in the Web cache performance evaluation and quantify their impact on the simulation accuracy. The results presented in this study show indeed some differences in the outcome of the simulation of a Web cache depending on the workload being used, and the probability distribution used to approximate updates on the cached documents. Each experiment shows two case studies that outline the impact of the considered parameter on the performance of the cache.
2010 IEEE Sixth International Conference on e-Science, 2010
The amount of computing resources currently available on Clouds is large and easily available wit... more The amount of computing resources currently available on Clouds is large and easily available with pay per use cost model. E-Science applications that need on-demand execution benefit from Clouds, because no permanent computing resources to support peak demand has to be acquired. In this paper, we present AMOS, a system that automates creation and management of temporary Grids on a
2013 IEEE 5th International Conference on Cloud Computing Technology and Science, 2013
ABSTRACT Cyber physical systems, such as intelligent dikes and smart energy systems, require scal... more ABSTRACT Cyber physical systems, such as intelligent dikes and smart energy systems, require scalable and flexible computing infrastructures to process data from instruments and sensor networks. Infrastructure as a Service clouds provide a flexible way to allocate remote distributed resources, but lack mechanisms to dynamically configure software (dependencies) and manage application execution. This paper describes the design and implementation of the Intercloud Operating System (ICOS), which acts between applications and distributed clouds, i.e., the Intercloud. ICOS schedules, configures, and executes applications in the Intercloud while taking data dependencies, budgets, and deadlines into account. Based on our experiences with the prototype, we present considerations and additional research challenges. The research on ICOS clarifies essential concepts needed to realize a flexible and scalable on-demand execution platform for distributed applications over distributed cloud providers.
Studies in health technology and informatics, 2006
Functional Magnetic Resonance Imaging (fMRI) is a popular tool used in neuroscience research to s... more Functional Magnetic Resonance Imaging (fMRI) is a popular tool used in neuroscience research to study brain activation due to motor or cognitive stimulation. In fMRI studies, large amounts of data are acquired, processed, compared, annotated, shared by many users and archived for future reference. As such, fMRI studies have characteristics of applications that can benefit from grid computation approaches, in which users associated with virtual organizations can share high performance and large capacity computational resources. In the Virtual Laboratory for e-Science (VL-e) Project, initial steps have been taken to build a grid-enabled infrastructure to facilitate data management and analysis for fMRI. This article presents our current efforts for the construction of this infrastructure. We start with a brief overview of fMRI, and proceed with an analysis of the existing problems from a data management perspective. A description of the proposed infrastructure is presented, and the cu...
Many e-Science applications are increasingly relying on orchestrating workflows of static web ser... more Many e-Science applications are increasingly relying on orchestrating workflows of static web services. The static nature of these web services means that workflow management systems have no control over the underlying mechanics of such services. This lack of control manifests itself as a problem when optimizing workflow execution since techniques such as data-locality aware deployment and service-to-service communication are very difficult to achieve. In this paper we propose a novel approach for mobilizing scientific web services onto common distributed resources and as such enable back-to-back communication between cooperating web services, autonomous web service scaling through fuzzy control and autonomous web service workflow orchestration.
Document coherency is a challenging problem for Web caching. Once the documents are cached throug... more Document coherency is a challenging problem for Web caching. Once the documents are cached throughout the Internet, it is often difficult to keep them coherent with the origin document without generating a new traffic that could increase the traffic on the international backbone and overload the popular servers. Several solutions have been proposed to solve this problem, among them two categories have been widely discussed: the strong document coherency and the weak document coherency. The cost and the efficiency of the two categories are still a controversial issue, while in some studies the strong coherency is far too expensive to be used in the Web context, in other studies it could be maintained at a low cost. The accuracy of these analysis is depending very much on how the document updating process is approximated. In this study, we compare some of the coherence methods proposed for Web caching. Among other points, we study the side effects of these methods on the Internet traffic. The ultimate goal is to study the cache behavior under several conditions, which will cover some of the factors that play an important role in the Web cache performance evaluation and quantify their impact on the simulation accuracy. The results presented in this study show indeed some differences in the outcome of the simulation of a Web cache depending on the workload being used, and the probability distribution used to approximate updates on the cached documents. Each experiment shows two case studies that outline the impact of the considered parameter on the performance of the cache.
2010 IEEE Sixth International Conference on e-Science, 2010
The amount of computing resources currently available on Clouds is large and easily available wit... more The amount of computing resources currently available on Clouds is large and easily available with pay per use cost model. E-Science applications that need on-demand execution benefit from Clouds, because no permanent computing resources to support peak demand has to be acquired. In this paper, we present AMOS, a system that automates creation and management of temporary Grids on a
2013 IEEE 5th International Conference on Cloud Computing Technology and Science, 2013
ABSTRACT Cyber physical systems, such as intelligent dikes and smart energy systems, require scal... more ABSTRACT Cyber physical systems, such as intelligent dikes and smart energy systems, require scalable and flexible computing infrastructures to process data from instruments and sensor networks. Infrastructure as a Service clouds provide a flexible way to allocate remote distributed resources, but lack mechanisms to dynamically configure software (dependencies) and manage application execution. This paper describes the design and implementation of the Intercloud Operating System (ICOS), which acts between applications and distributed clouds, i.e., the Intercloud. ICOS schedules, configures, and executes applications in the Intercloud while taking data dependencies, budgets, and deadlines into account. Based on our experiences with the prototype, we present considerations and additional research challenges. The research on ICOS clarifies essential concepts needed to realize a flexible and scalable on-demand execution platform for distributed applications over distributed cloud providers.
Studies in health technology and informatics, 2006
Functional Magnetic Resonance Imaging (fMRI) is a popular tool used in neuroscience research to s... more Functional Magnetic Resonance Imaging (fMRI) is a popular tool used in neuroscience research to study brain activation due to motor or cognitive stimulation. In fMRI studies, large amounts of data are acquired, processed, compared, annotated, shared by many users and archived for future reference. As such, fMRI studies have characteristics of applications that can benefit from grid computation approaches, in which users associated with virtual organizations can share high performance and large capacity computational resources. In the Virtual Laboratory for e-Science (VL-e) Project, initial steps have been taken to build a grid-enabled infrastructure to facilitate data management and analysis for fMRI. This article presents our current efforts for the construction of this infrastructure. We start with a brief overview of fMRI, and proceed with an analysis of the existing problems from a data management perspective. A description of the proposed infrastructure is presented, and the cu...
A scientific workflow management system can be considered as a binding agent which brings togethe... more A scientific workflow management system can be considered as a binding agent which brings together scientists and distributed resources. A workflow graph plays the central role in such a system as it is the component understood by both scientist and machine. Making sense out of a scientific workflow graphs is undoubtedly, the first and foremost responsibility of a workflow management system. Typical systems include an orchestration engine which models a workflow and schedules individual components onto distributed resources. As part of the WS-VLAM, we present an alternative orchestration engine which takes a different stand on interpreting the workflow graph. Whilst the current engine in WS-VLAM models the graph as a process network where components are tightly coupled through communication channels, the Datafluo engine models the graph as a dataflow network with farming capabilities. In this dissertation, we present the Datafluo architecture followed by a prototype implementation. The prototype is taken through its passes using scientific workflow applications where the generated results demonstrate the orchestration features. Through our results, we show how dataflow techniques reduce queue waiting times whilst farming techniques circumvent common workflow bottlenecks.
Uploads
Papers by Adam Belloum
systems include an orchestration engine which models a workflow and schedules individual components onto distributed resources. As part of the WS-VLAM, we
present an alternative orchestration engine which takes a different stand on interpreting the workflow graph. Whilst the current engine in WS-VLAM models the graph as a process network where components are tightly coupled through communication channels, the Datafluo engine models the graph as a dataflow network with farming capabilities. In this dissertation, we present the Datafluo architecture followed by a prototype implementation. The prototype is taken through its passes using scientific workflow applications where the generated results demonstrate the orchestration features. Through our results, we show how dataflow techniques reduce queue waiting times whilst farming techniques circumvent common workflow bottlenecks.