V.N. Vasilev, K.V. Knyazkov, T.N. Tchurov, D.A. Nasonov, S.V. Maryin,
S.V. Kovalchuk, A.V. Boukhanovsky
The article describes technological CLAVIRE platform capabilities for second generation cloud computing environment building for processing of large data volumes. Peculiarities of contemporary tasks for information and control systems development within Data-Driven Approach (DDA) are concerned. Cloud computing is used as a core concept for resource providing access to the software and hardware resources. Cloud Computing Maturity Model is concerned as an extension which defines the set of key requirements to cloud computing environments. Number of technologies is implemented within CLAVIER to answer full functional requirements of cloud computing environment.
The architecture of CLAVIRE platform is defined by the following key characteristics:
• Domain-specific language (EasyPackage) is used to describe available software packages and data sources, their calling procedure, parameter passing and used data formats.
• Abstract workflow description using domain specific language (EasyFlow) is used to define a composition application to be run.
• Interpretation of the domain-specific languages produces composite application, which calls computational and data services within cloud environment.
• Dynamic scheduling of the composite application allows overcoming performance issues using parameterized models for prediction of computation time.
• Different computational resources are integrated within the single environment and can be called using unified access interface.
• High-level graphical user interfaces are presented for human–computer interaction within problem domain.
• Knowledge processing allows to present the intelligent user support during whole process of simulation and data processing.
• Dynamic access to data sources and visualization facilities allows to ais provided by Scientific Visualization service (may be interpreted as local software).
Five classes of user interfaces for access to the platform within different levels of abstraction are described: console and programming interface, problem-oriented interface, workflow management system interface, intelligent advisor interface, Virtual Simulation Objects interface. Mentioned classes of interfaces allow to provide the user with appropriate level of service support according to the requirements of cloud computing.
Example composite application being built using CLAVIRE platform within DDA is concerned: composite application for forecasting of critical situations with terrorists attack during unauthorized flash-mob meeting.