Reproducible Workﬂow on a Public Cloud for Computational Fluid Dynamics - Part 1 - Reproducible research and cloud benchmarks

Oct 16, 2026

The paper presents a reproducible workflow for computational studies executed and shared on Microsoft Azure, addressing how to provide transparency for evidence-based science. Docker containers are used to package the application software stack, while additional tools automate virtual machine creation and job submission. Configuration files are included in expanded reproducibility packages with workflow definitions, input files, and instructions, enabling others to recreate the cloud environment and rerun computations under identical conditions. Benchmarks for tightly coupled applications test whether Azure HPC nodes can replace traditional on-site clusters, and GPU-parallel research software is used to complete computational fluid dynamics studies.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Introduction

Definitions and motivation for reproducible research

Container-based reproducibility workflow on Azure

Benchmarks and viability of Azure HPC nodes

Reproducible Workﬂow on a Public Cloud for Computational Fluid Dynamics arXiv : 1904 .07981v3 [ cs .CE] 25 Sep 2019 Olivier Mesnard, Lorena A. Barba Abstract—In a new effort to make our research transparent and reproducible by others, we developed a workﬂow to run and share computational studies on the public cloud Microsoft Azure. It uses Docker containers to create an image of the application software stack. We also adopt several tools that facilitate creating and managing virtual machines on compute nodes and submitting jobs to these nodes. The conﬁguration ﬁles for these tools are part of an expanded “reproducibility package” that includes workﬂow deﬁnitions for cloud computing, in addition to input ﬁles and instructions. This facilitates re-creating the cloud environment to re-run the computations under the same conditions. Although cloud providers have improved their offerings, many researchers using high-performance computing (HPC) are still skeptical about cloud computing. Thus, we ran benchmarks for tightly coupled applications to conﬁrm that the latest HPC nodes of Microsoft Azure are indeed a viable alternative to traditional on-site HPC clusters. We also show that cloud offerings are now adequate to complete computational ﬂuid dynamics studies with in-house research software that uses parallel computing with GPUs. Finally, we share with the community what we have learned from nearly two years of using Azure cloud to enhance transparency and reproducibility in our computational simulations. ~~ ~~ F ~~ ~~ 1 INTRODUCTION REPRODUCIBLE research and replication studies are es sential components for the progress of evidence-based science, even more now when nearly all ﬁelds advance via computation. We use computer simulations and data models to create new knowledge, but how do we provide evidence that this new knowledge is justiﬁed? Traditional journal publications exclude software and data products from the peer-review process, yet reliance on ever more complex computational artifacts and methods is the norm. Lacking standards for documenting, reporting and reviewing the computational facets of research, it becomes difﬁcult to verify and corroborate the ﬁndings presented in journalsand conferences [1] . The literature is cluttered with confused and sometimes contradictory deﬁnitions for reproducible research, reproducibility, replicability, repetition, etc. [2] . It is thus worth clarifying how we use these terms. “Reproducible research” was used by geophysics professor Jon Claerbout in the 1990s to mean computational studies that are published with sufﬁcient transparency so other scientists can recreate the results. His research group at Stanford created a reproducible-research environment [3] whose goal was complete documentation of scientiﬁc computations, in such away that a reader could reproduce all the results and ﬁguresin a paper using the author-provided computer programs and raw data. This requires open data and open source software and, for this reason, the reproducibility movement is closely linked with the open science movement. The term“replication” has been adopted to refer to an independent study generating new data which, when analyzed, lead to the same ﬁndings [4] . We follow this convention, adopted also recently in reports from the National Academies of 􀀏 Mechanical and Aerospace Engineering, the George Washington University, Washington, DC 20052. [E-mail: mesnardo@gwu.edu](E-mail: mesnardo@gwu.edu) 􀀏 Email: [labarba@gwu.edu](labarba@gwu.edu) Sciences, Engineering, and Medicine [5],[6] . Many efforts to develop cyberinfrastructure that supports reproducible research have been launched in past years. They address concerns like automatic capture of changes to software (version control systems), persistent data archival, global registration of data identiﬁers, workﬂow management, and more. But capturing the whole computational environment used in a research project remains one

FAQs

What workflow is proposed for reproducible computational fluid dynamics studies?

The workflow enables running and sharing computational studies on Microsoft Azure using Docker containers plus tools for virtual machines and job submission, with configuration files packaged for full recreation of the environment.

How does the paper clarify terms like reproducible research and replication?

It explains reproducible research as computational studies published with sufficient transparency for others to recreate results, and uses replication to mean an independent study generating new data that leads to the same findings.

Why do the authors benchmark Microsoft Azure for HPC workloads?

They benchmark tightly coupled applications to verify that the latest Azure HPC nodes are a viable alternative to traditional on-site HPC clusters, especially given that many researchers remain skeptical about cloud computing.

Oct 16, 2026

Uploaded by Mafia Boss