F. Guim

Publications (11)0 Total impact

• Source

  Available from: Ivan Rodero

  Conference Proceeding: A google earth based distributed infrastructure to support natural disaster response

  F. Guim, I. Rodero, E. Ozalp, S. Madry
  [show abstract] [hide abstract]
  ABSTRACT: Forecasting and monitoring earth phenomenon is fundamental to reducing natural disasters hazards. However, how to react after a given disaster occurs is also a key issue that can improve response times and save lives. GISCorps is a program of the urban and Regional Information Systems Association (URISA) that coordinates short-term, volunteer-based GIS services to underprivileged communities. In case of a natural disaster, the GISCorps members volunteer their time to characterize the prior and current status of the affected area (for example doing damage assessment) by the observation of satellites images and ancillary data using Google Earth. This information is crucial since it can help local services carry out better decisions and for scheduling response strategies. This methodology has the main advantage of being simple, easy to schedule and based in uniform and standard formats. However its main drawback is that it lacks scalability and relies only in the level of expertise of the member and his/her knowledge and tools. Current high performance computing and information retrieval techniques can provide to this community powerful features that can speed up and improve their procedures. Furthermore, they can provide value added mechanisms to automatically collect and process data concerning the disaster. This new knowledge can be used by the GIS expert to better identify areas or can be automatically forwarded to the local services. In this paper we propose a grid based new service architecture oriented to improved the GISCorps procedures and improve risk reduction in natural disasters. The main aim of this paper is to present a use case of how the forthcoming technologies can be used in natural disaster scenarios to improve response time and provide better information for the decision making process.
  Recent Advances in Space Technologies, 2009. RAST '09. 4th International Conference on; 07/2009
• Source

  Available from: Ivan Rodero

  Conference Proceeding: A novel framework for a unified international system of volcano early warning and hazard tracking

  F. Guim, I. Rodero, E. Ozalp, E. Chester
  [show abstract] [hide abstract]
  ABSTRACT: While numerous global initiatives exist to address the potential hazards posed by volcanic eruption events and assess impacts from a civil security viewpoint, there is still no single, unified, international system of early warning and hazard tracking for eruptions. Numerous gaps exist in the risk reduction cycle, from data collection, to data processing, and finally dissemination of salient information to relevant parties. As part of the International Space University's Space Studies Program during summer 2008, a detailed gap analysis of the state of volcano disaster risk reduction was undertaken. Using this analysis as a basis, a novel framework to deal with all the identified gaps is proposed as an outcome of this study. This paper presents the principal innovations concerning the data and information processing within this work. The name of the framework was chosen to be VIDA to abbreviate dasiaVolcanic Activity Processing of Observation and Remote Sensing Data Integrated Data-Sharing and Analysispsila. In order to present this framework, it is shown how different high performance computing resources can be orchestrated in order to support fast and efficient tools that would support VIDA in mitigating the risk posed by volcanic eruptions and other natural hazards. Furthermore, how the system could be used by describing different use cases is also presented in this paper.
  Recent Advances in Space Technologies, 2009. RAST '09. 4th International Conference on; 07/2009
• Source

  Available from: upc.edu

  Conference Proceeding: Prediction f based models for evaluating backfilling scheduling policies

  F. Guim, J. Corbalan, J. Labarta
  [show abstract] [hide abstract]
  ABSTRACT: The research on the usage of prediction techniques in HPC scheduling policies rather than user estimates has increased it relevance these recent years. In the coming scheduling architectures, like grids and very heterogeneous computational resources, such techniques are having a crucial relevance due to users in most of the cases will not have enough information or enough skills for specify for how long will their jobs run. Many studies have analyzed the impact of the user runtime estimates accuracy in the performance of the scheduling policies. Using user runtime estimation models, such as the f-model, researchers have evaluated how the accuracy of the runtime estimates provided by the user at the job submission can affect the performance of the backfilling policies and its variants. However, these traditional estimation models can not applied to backfilling scheduling policies that use runtime predictions rather than user estimates. Clearly, predictions can not be characterized with these models. For instance because the underestimation of the runtime is not considered by them and obviously it can occurs. In this paper we describe and evaluate a set of f-model based prediction models that characterize the behavior that prediction techniques have shown in HPC centers. They have been designed for evaluate scheduling policies that use predictions rather than user estimates.
  Parallel and Distributed Computing, Applications and Technologies, 2007. PDCAT '07. Eighth International Conference on; 01/2008
• Article: The Palantir Grid Meta-Information System

  F. Guim, I. Rodero, M. Tomas, J. Corbalan, J. Labarta
  [show abstract] [hide abstract]
  ABSTRACT: Grids allow large scale resource-sharing across different administrative domains. Those diverse resources are likely to join or quit the grid at any moment or possibly to break down. Grid monitoring tools have to adapt supporting access information to these heterogeneous and not reliable environments. There is a wide rage of types of resources to be monitored, with different nature, characteristics and so on. For instance grid users want to monitor from which is the state of the jobs that submit till how many processors has a given resources of a given centre. These issues make the task of monitoring complex to treat, and it is difficult to provide a generals ways for accessing to all this information. In this work we propose a set of functionalities that a grid information system should provide. We describe the Palantir meta-information system that has been designed for uniform the access to different monitoring and information systems and that implements all the discussed functionalities. This meta-information system tries to abstract the resource information that underlies on the local systems, and homogenize the access to all this. We present the architecture of Palantir and its main characteristics
  Grid Computing, IEEE/ACM International Workshop on. 09/2006;
• Source

  Available from: Ivan Rodero

  Conference Proceeding: Uniform Job Monitoring using the HPC-Europa Single Point of Access

  F. Guim, I. Rodero, J. Corbalan, J. Labarta, A. Oleksiak, T. Kuczynski, D. Szejnfeld, J. Nabrzyski
  [show abstract] [hide abstract]
  ABSTRACT: Job monitoring in Grid systems presents an important challenge due to Grid environments are volatile, heterogeneous, not reliable and are managed by different middlewares and monitoring tools. We present the infrastructure that we have designed and implemented in the HPC-Europa European project that allows uniform access to job monitoring information from different virtual organizations. The presented system abstracts user to the complexities of the underlying systems of each middleware. The API that each center has to implement for providing access to its job monitoring information is explained. Finally, we show all the features that user can use in the portal to personalize his/her monitoring environment: how and which information has to be presented.
  Cluster Computing and the Grid Workshops, 2006. Sixth IEEE International Symposium on; 06/2006
• Source

  Available from: Ivan Rodero

  Conference Proceeding: How the JSDL can exploit the parallelism?

  I. Rodero, F. Guim, J. Corbalan, J. Labarta
  [show abstract] [hide abstract]
  ABSTRACT: The description of the jobs is a very important issue for the scheduling and management of grid jobs. Since there are a lot of different languages for describing grid jobs, the GGF have presented the Job Submission Description Language (JSDL) to standardize the job submission language. We believe that the JSDL is a good solution but it has some deficiencies regarding the parallelism issues. In this paper, we propose an extension of the JSDL to specify the parallelism details of grid jobs. This extension is proposed in general terms for supporting current multilevel parallel applications and incoming approaches in parallel programming models. We also discus the suitability of the multilevel parallel programming models for grids, in particular the MPI+OpenMP since our project, the eNANOS project, is based on this hybrid programming model.
  Cluster Computing and the Grid, 2006. CCGRID 06. Sixth IEEE International Symposium on; 06/2006
• Source

  Available from: upc.edu

  Article: Impact of qualitative and quantitative errors of the job runtime estimation in backfilling based scheduling policies

  F Guim, J Corbalan, J Labarta
  [show abstract] [hide abstract]
  ABSTRACT: Estimation or prediction accuracy in backfilling based poli-cies it's an important issue that has high impact on the system performance. However it's not clear which is the required pre-cision of such estimations, moreover it's not clear which kind of errors are critical when scheduling the jobs that are queued in the local systems. In this paper we present a deeper anal-ysis of the impact of the estimation errors in the scheduling. The study is based on several criteria. For instance the im-plications of carrying out accurate predictions of determined kind of jobs, having qualitative errors or quantitative errors and so on. As far as we know, all the works that have analyzed the im-pact of the estimation error in the backfilling policies based they results and conclusions in a set of well known metrics, such as the average slowdown or the wait time. We present a new parametrical metric that evaluates the index of satisfac-tion of the user. As shown in the paper, conclusions obtained when evaluating backfilling based policies with this metric may differ with other conclusions obtained, for example, with average slowdown.
• Source

  Available from: psu.edu

  Article: A job self-scheduling policy for HPC infrastructures

  F Guim, J Corbalan
  [show abstract] [hide abstract]
  ABSTRACT: The number of distributed high performance computing architectures has increased exponentially these last years. Thus, systems composed by several computational resources provided by different Research centers and Universities have become very popular. Job scheduling policies have been adapted to these new scenarios in which several independent resources have to be managed. New policies have been designed to take into account issues like multi-cluster environ-ments, heterogeneous systems and the geographical distribution of the resources. Several centralized scheduling solutions have been proposed in the literature for these environments, such as centralized schedulers, centralized queues and global controllers. These approaches use a unique scheduling entity responsible for scheduling all the jobs that are submitted by the users. In this paper we propose the usage of self-scheduling techniques for dispatching the jobs that are submitted to a set of distributed computational hosts that are managed by independent schedulers (such as MOAB or LoadLeveler). It is a non-centralized and job-guided scheduling policy whose main goal is to optimize the job wait time. Thus, the scheduling decisions are done independently for each job instead of using a global policy where all the jobs are considered. On top of this, as a part of the proposed solution, we also demonstrate how the usage of job wait time prediction techniques can substantially improve the performance obtained in the described architecture.
• Article: The XtreemOS jScheduler: using self-scheduling techniques in large computing architectures

  F. Guim, I. Rodero, M. Garcia, J. Corbalan
  [show abstract] [hide abstract]
  ABSTRACT: Several centralized scheduling solutions have been proposed in the literature for environments composed of several independent computational resources, such as centralized schedulers, centralized queues and global controllers. These approaches use a unique scheduling entity responsible for scheduling all jobs submitted by users. In our previous work we proposed the use of self-scheduling techniques to dispatch jobs which are submitted to a set of distributed computational hosts, which are in turn managed by independent schedulers (such as MOAB or LoadLeveler). In the ISIS-Dispatcher, scheduling decisions are made independently for each job instead of using a global policy where all jobs are considered. In this paper we present how the ISIS-Dispatcher techniques can be used in the XtreemOS architecture for manage the jobs. This system is designed to be deployed in large scenarios that potentially involve thousands of resources. In such systems it is not feasible to make the dispatcher contact to all the systems. It is not realistic to suppose that the dispatcher stores the information about all the resources and where they are located. Obviously, this approach would imply problems of scalability. In this paper we also evaluate the impact about the amount of resource information that the dispatcher can collect during the job scheduling.
• Source

  Available from: Ivan Rodero

  Article: Portal de acceso a recursos HPC en entornos Grid

  F. Guim, I. Rodero, J. Corbalán, J. Labarta
  [show abstract] [hide abstract]
  ABSTRACT: En un entorno Grid formado por diversos centros computacionales con diversas tecnologías e infraestructuras involucradas, resulta muy útil disponer de sistemas para proporcionar un único punto de acceso a los recursos de forma sencilla y transparente. Dentro del proyecto HPC-Europa se ha desarrollado un sistema de acceso a diversos centros europeos con tradición en computación de altas prestaciones basado en un portal web. En este texto se presentan las principales características de este sistema y se describe una demostración efectuada de su funcionamiento realizada durante la tercera reunión de la Red Temática en Grid Middleware.
  RedIRIS: boletín de la Red Nacional de I+D RedIRIS, ISSN 1139-207X, Nº. 80, 2007, pags. 67-71.
• Source

  Available from: Ivan Rodero

  Article: Looking for an evolution of grid scheduling: Meta-brokering

  I Rodero, F Guim, J Corbalan, L L Fong, Y G Liu, S M Sadjadi
  [show abstract] [hide abstract]
  ABSTRACT: A Grid Resource Broker for a Grid domain, or also known as meta-scheduler, is a middleware component used for matching works to available Grid resources from one or more IT organizations. A Grid meta-scheduler usually has its own interfaces for the functionalities it provides and its own job scheduling objectives. This situation causes two main problems: the user uniform access to the Grid is lost, and the scheduling decisions are taken separately while they should be done in coordination. These problems have been observed in different efforts such as the HPC-Europa project but they are still open problems. In this paper we discuss the requirements to achieve a more uniform access to the Grids through a new approach to global brokering with new scheduling techniques. As the results of these discussions on brokering requirements, we propose a meta-brokering design, so called meta-meta-scheduler design, and discuss how it can be realized as a centralized model for the HPC-Europa project, and as a distributed model for the LA Grid project.