A5017532452- Cloud Computing and Resource Management
- Distributed and Parallel Computing Systems
- Scientific Computing and Data Management
- IoT and Edge/Fog Computing
- Advanced Data Storage Technologies
- Parallel Computing and Optimization Techniques
- Software System Performance and Reliability
- Distributed systems and fault tolerance
- Data Visualization and Analytics
- Peer-to-Peer Network Technologies
- Data Stream Mining Techniques
- Cloud Data Security Solutions
- Computer Graphics and Visualization Techniques
- Caching and Content Delivery
- Meteorological Phenomena and Simulations
- Remote Sensing in Agriculture
- Seismic Imaging and Inversion Techniques
- Climate change impacts on agriculture
- Advanced Queuing Theory Analysis
- Personal Information Management and User Behavior
- Smart Agriculture and AI
- Mobile Agent-Based Network Management
- Advanced Image Processing Techniques
- Big Data and Business Intelligence
- Software-Defined Networks and 5G
Microsoft (United States)
2024
IBM (Brazil)
2020
IBM Research - Brazil
2012-2019
IBM (United States)
2011-2018
The University of Melbourne
2006-2013
Interface (United Kingdom)
2013
Pontifícia Universidade Católica do Rio Grande do Sul
2003-2012
Cloud Computing Center
2011
Universidade de São Paulo
2005-2006
Universidade Estadual de Campinas (UNICAMP)
2002
SUMMARY Cloud computing allows the deployment and delivery of application services for users worldwide. Software as a Service providers with limited upfront budget can take advantage lease required capacity in pay‐as‐you‐go basis, which also enables flexible dynamic resource allocation according to service demand. One key challenge potential customers have before renting resources is know how their will behave set costs involved when growing shrinking pool. Most studies this area rely on...
Running high-resolution physical models is computationally expensive and essential for many disciplines. Agriculture, transportation, energy are sectors that depend on weather models, which typically consume hours of large High Performance Computing (HPC) systems to deliver timely results. Many users cannot afford run the desired resolution forced use low output. One simple solution interpolate results visualization. It also possible combine an ensemble obtain a better prediction. However,...
Cloud resources and services are offered based on Service Level Agreements (SLAs) that state usage terms penalties in case of violations. Although, there is a large body work the area SLA provisioning monitoring at infrastructure platform layers, SLAs usually assumed to be guaranteed application layer. However, challenging task due monitored metrics or layer cannot easily mapped required Sophisticated among those layers avoid costly maximize provider profit still an open research challenge....
Auto-scaling is a key feature in clouds responsible for adjusting the number of available resources to meet service demand. Resource pool modifications are necessary keep performance indicators, such as utilisation level, between user-defined lower and upper bounds. strategies that not properly configured according user workload characteristics may lead unacceptable QoS large resource waste. As consequence, there need deeper understanding auto-scaling how they should be minimise these...
Yield forecast is essential to agriculture stakeholders and can be obtained with the use of machine learning models data coming from multiple sources. Most solutions for yield rely on NDVI (Normalized Difference Vegetation Index) which, besides being time-consuming acquire process, only allows forecasting once crop season has already started. To bring scalability forecast, in present paper we describe a system that incorporates satellite-derived precipitation soil properties datasets,...
The Cloud computing paradigm has revolutionised the computer science horizon during past decade and enabled emergence of as fifth utility. It captured significant attention academia, industries, government bodies. Now, it emerged backbone modern economy by offering subscription-based services anytime, anywhere following a pay-as-you-go model. This instigated (1) shorter establishment times for start-ups, (2) creation scalable global enterprise applications, (3) better cost-to-value...
Machine Learning (ML) has become essential in several industries. In Computational Science and Engineering (CSE), the complexity of ML lifecycle comes from large variety data, scientists' expertise, tools, workflows. If data are not tracked properly during lifecycle, it becomes unfeasible to recreate a model scratch or explain stackholders how was created. The main limitation provenance tracking solutions is that they cannot cope with capture integration domain processed multiple workflows...
Metaschedulers can distribute parts of a bag-of-tasks (BoT) application among various resource providers in order to speed up its execution. When cannot disclose private information such as their load and computing power, which are usually heterogeneous, the metascheduler needs make blind scheduling decisions. We propose three policies for composing offers schedule deadline-constrained BoT applications. Offers act mechanism expose interest executing an entire or only part it without...
Resource allocation in High Performance Computing (HPC) settings is still not easy for end-users due to the wide variety of application and environment configuration options. Users have difficulties estimate number processors amount memory required by their jobs, select queue partition, when job output will be available plan next experiments. Apart from wasting infrastructure resources making wrong decisions, overall user response time can also negatively impacted. Techniques that exploit...
The more instrumented society is demanding smarter services to help coordinate daily activities and exceptional situations. Applications become sophisticated context-aware as the pervasiveness of technology increases. In order cope with resource limitations mobile-based environments, it a common practice delegate processing intensive components Cloud Computing infrastructure. this scenario, executions server-based jobs are still dependent on local variations end-user context. We claim that...
Large-scale computing environments, such as TeraGrid, Distributed ASCI Supercomputer (DAS), and Gridpsila5000, have been using resource co-allocation to execute applications on multiple sites. Their schedulers work with requests that contain imprecise estimations provided by users. This lack of accuracy generates fragments inside the scheduling queues can be filled rescheduling both local multi-site requests. Current solutions rely advance reservations ensure users access all resources at...
Resource allocation in High Performance Computing (HPC) settings is still not easy for end-users due to the wide variety of application and environment configuration options. Users have difficulties estimate number processors amount memory required by their jobs, select queue partition, when job output will be available plan next experiments. Apart from wasting infrastructure resources making wrong decisions, overall user response time can also negatively impacted. Techniques that exploit...
Computational Science and Engineering (CSE) projects are typically developed by multidisciplinary teams. Despite being part of the same project, each team manages its own workflows, using specific execution environments data processing tools. Analyzing processed all workflows globally is a core task in CSE project. However, this analysis hard because generated these not integrated. In addition, since may take long time to execute, needs be done at runtime reduce cost A typical solution...