%0 Conference Paper
%B Computer Modeling and Intelligent Systems CMIS-2020
%D 2020
%T Docker Container based PaaS Cloud Computing Comprehensive Benchmarks using LAPACK
%A Dmitry Zaitsev
%A Piotr Luszczek
%K docker containers
%K software containers
%X Platform as a Service (PaaS) cloud computing model becomes wide- spread implemented within Docker Containers. Docker uses operating system level virtualization to deliver software in packages called containers. Containers are isolated from one another and comprise all the required software, including operating system API, libraries and configuration files. With such advantageous integrity one can doubt on Docker performance. The present paper applies packet LAPACK, which is widely used for performance benchmarks of super- computers, to collect and compare benchmarks of Docker on Linux Ubuntu and MS Windows platforms. After a brief overview of Docker and LAPACK, a se- ries of Docker images containing LAPACK is created and run, abundant benchmarks obtained and represented in tabular and graphical form. From the final discussion, we conclude that Docker runs with nearly the same perfor- mance on both Linux and Windows platforms, the slowdown does not exceed some ten percent. Though Docker performance in Windows is essentially lim- ited by the amount of RAM allocated to Docker Engine.
%B Computer Modeling and Intelligent Systems CMIS-2020
%C Zaporizhzhoa
%8 2020-03
%G eng
%0 Journal Article
%J IEEE Transactions on Parallel and Distributed Systems
%D 2019
%T Solving Linear Diophantine Systems on Parallel Architectures
%A Dmitry Zaitsev
%A Stanimire Tomov
%A Jack Dongarra
%K Mathematical model
%K Matrix decomposition
%K Parallel architectures
%K Petri nets
%K Software algorithms
%K Sparse matrices
%K Task analysis
%X Solving linear Diophantine systems of equations is applied in discrete-event systems, model checking, formal languages and automata, logic programming, cryptography, networking, signal processing, and chemistry. For modeling discrete systems with Petri nets, a solution in non-negative integer numbers is required, which represents an intractable problem. For this reason, solving such kinds of tasks with significant speedup is highly appreciated. In this paper we design a new solver of linear Diophantine systems based on the parallel-sequential composition of the system clans. The solver is studied and implemented to run on parallel architectures using a two-level parallelization concept based on MPI and OpenMP. A decomposable system is usually represented by a sparse matrix; a minimal clan size of the decomposition restricts the granulation of the technique. MPI is applied for solving systems for clans using a parallel-sequential composition on distributed-memory computing nodes, while OpenMP is applied in solving a single indecomposable system on a single node using multiple cores. A dynamic task-dispatching subsystem is developed for distributing systems on nodes in the process of compositional solution. Computational speedups are obtained on a series of test examples, e.g., illustrating that the best value constitutes up to 45 times speedup obtained on 5 nodes with 20 cores each.
%B IEEE Transactions on Parallel and Distributed Systems
%V 30
%P 1158-1169
%8 2019-05
%G eng
%U https://ieeexplore.ieee.org/document/8482295
%N 5
%R http://dx.doi.org/10.1109/TPDS.2018.2873354