Ruslan Pacevič

Development of cloud services for patient-specific simulations of blood flows through aortic valves

The paper presents the development of cloud software services for patient-specific computational analysis of blood flows through the aortic valve on a private university cloud. The main focus is on the software service level at the top of the provided computational platform. Blood flow through the aortic valve was considered as a pilot application of the OpenStack cloud infrastructure. A modelling software environment based on ANSYS Fluent was developed as a software service (SaaS) for the numerical analysis of low flow, low pressure gradient aortic stenosis. Segmentation software services were designed to deal with the patient-specific issues of the computational analysis. User-friendly management tools were developed using Apache jclouds API to enhance the management of OpenStack cloud infrastructure and to increase the accessibility of the required software. The performance of the cloud infrastructure was assessed by testing CPU, memory bandwidth, disk I/O and the developed software service for medical computations. The performance measured on Xen hardware virtual machines, KVM virtual machines and Docker containers were compared with the performance obtained by using the native hardware.

VizLitG: Grid Visualization e-Service Enabling Partial Dataset Transfer from Storage Elements of gLite-based Grid Infrastructure

Grid visualization e-service VizlitG is developed for Lithuanian grid infrastructure based on gLite middleware. The client–server architecture of the e-service VizLitG is based on widely recognized web standards implemented in Java EE 6 platform and GlassFish application server. VizLitG is designed for convenient access and interactive visualization of remote data files located in storage elements of gLite-based grid infrastructure. Partial dataset transfer from the experimental storage element is developed in order to reduce the communication time and achieve interactive rates. The performance of VizLitG is evaluated visualizing poly-dispersed particle systems of different sizes in a geographically distributed grid. The attention is focused on the performance of data transfer from the storage elements and its contribution to the total benchmark time. Different types of software are employed for data transfer in order to present the quantitative comparison.

Visualization of cracks by using the local Voronoi decompositions and distributed software

Local Voronoi decompositions are applied to extraction of crack surfaces.Cracks propagate in mono-dispersed particulate media simulated by the DEM.The geometry of crack surfaces is accurately defined by faces of the decomposition.Generation of local decompositions rather than global meshes reduces execution time.The sufficiently high speed-up of distributed visualization software is achieved. The paper presents a novel visualization technique for cracks propagating in mono-dispersed particulate material. The proposed technique is based on local space decompositions generated in fractured areas. The contact surfaces of the neighboring particles are defined by the local Voronoi decomposition generated according to the lattice topology employed in computations of the discrete element method. The visual model validation helps to indicate the regions of a highly deformed lattice, where the defects detected between the pairs of the neighboring particles on the lattice connections cannot be directly mapped onto the relevant edges of the Voronoi diagram. The parallel implementation of the visualization technique is based on the domain decomposition and two layers of ghost vertices and connections. The technique is implemented in the distributed visualization software VisPartDEM. Datasets of the elastic solid problem exhibiting non-uniform distribution of fracture force values are considered to validate the performance of the proposed technique. The parallel speed-up of the visualization software is investigated. The superior performance of the applied local technique is compared to the performance observed by using the standard global Voronoi algorithm.

On Efficiency of Parallel Solvers for the Blood Flow through Aortic Valve

Mathematical modelling of cardiac haemodynamics presents a great challenge to the computational scientists due to numerous numerical issues and required computational resources. In this paper, we study the parallel performance of 3D simulation software for the blood flow through the aortic valve. The fluid flow problem with the open aortic valve leaflets is formulated and solved in parallel. The choice between the segregated and coupled numerical schemes is discussed and investigated. We present and compare the parallel performance results of both types of parallel solvers. We investigate their strong and weak scalability.

The development of VisLT visualization service in Openstack cloud infrastructure

VisLT visualization service is developed in the OpenStack cloud infrastructure.The partial dataset transfer from the object storage is developed as the middleware.The partial dataset transfer took from 1.7% to 6.7% of the full dataset transfer.Virtualized GPU rendering and monitoring is implemented on Xen HVM.The performance difference between the native hardware and the VM is equal to 5.6%. The paper presents the development of visualization software as a service in the OpenStack cloud infrastructure. VisLT cloud visualization service is developed for visualizing the results computed and stored in the private cloud infrastructure. Partial dataset transfer from the OpenStack object storage is developed to reduce the communication between different parts of the cloud infrastructure. The cloud services are developed by using Apache jclouds API to enhance the management of cloud infrastructure and to increase the accessibility of the visualization service. GPU is virtualized as a PCI device employing direct pass-through technology on the hardware virtual machines of Xen hypervisor to ensure fast remote rendering, which is a key feature of distributed visualization systems. The extraction of crack surfaces and computations of flows through the aortic valve are considered as pilot applications of the developed visualization service. The performance of VisLT measured on the XEN hardware virtual machines is compared to the performance attained on the native hardware.

Private Cloud Infrastructure for Applications of Mechanical and Medical Engineering

The paper presents the development of the private cloud infrastructure providing advanced services for research in mechanical and medical engineering. The cloud services have been developed by using Apache jclouds API and JetS3t Toolkit to enhance management of Eucalyptus cloud infrastructure and to increase accessibility of engineering software. The performance of the developed cloud infrastructure has been assessed testing CPU, memory IO, disk IO, network and the deployed software services. The obtained results have been compared with the performance of the native hardware. The technical implementation of deployed cloud services has been evaluated by the utilising a number of use cases from different research areas of mechanical and medical engineering. DOI: http://dx.doi.org/10.5755/j01.itc.44.3.7379

Cell Attribute-Based Algorithm for Crack Visualization

The paper presents the development of the visualization algorithm for propagating cracks. The novel algorithm is based on the cell attribute obtained from the damaged lattice connections employed for discrete element computations of mono-dispersed particulate media. Generation of the cells is efficiently performed by using the positions of particles and the lattice connections. The developed visualization algorithm is implemented in the distributed visualization software VisPartDEM. The efficiency of the software is tested visualizing the datasets resulting from computations of the lattice-based discrete element method. The performance of the developed algorithm is compared with that of the visualization algorithms based on the Voronoi diagrams and the inscribed spheres. DOI: http://dx.doi.org/10.5755/j01.itc.42.3.2575

Influence of Network Communications to the Final Performance of Grid Visualization Software

Influence of network communications to the final performance of grid visualization software VisPartDEM is investigated. Distributed architecture of VisPartDEM is designed for interactive visualization of large particle systems simulated by the discrete element method on the gLite-based grid. Performance analysis of VisPartDEM is made on the geographically distributed gLite grid infrastructure. The attention is focused on the data transfer from storage elements, image transfer to clients and their contribution to the total benchmark time. The discussed issues are important for all researchers interactively visualizing results on gLite-based remote grid infrastructures.