A. Feoktistov

Effective solving complex mathematical modeling problems is based on the use of high-performance computing. Clouds, grids, and public access supercomputer centers are commonly used platforms. Their integration into a unified environment provides possibilities for carrying out mass large-scale scientific experiments and efficient scalable resource allocation at different stages of the application design and execution. However, end-users have to carefully select optimization criteria such as completion time, deadlines, reliability, cost, etc. It is a complicated problem due to integrated resources differ significantly in their computing capabilities, hardware and software platforms, system architectures, user interfaces, etc. The paper presents new features of the Orlando Tools framework for the development of distributed applied software packages (scalable scientific applications) that mitigates various types of uncertainties arising from the job distribution in the integrated computing environment. It provides continuous integration, delivery, and deployment of applied and system software to significantly mitigate the negative impact of uncertainty on problem-solving time, computation reliability, and resource efficiency. An experimental analysis of the sustainable design and development of the real energy sector clearly demonstrates the advantages of the tools.