Skyport - Container-Based Execution Environment Management for Multi-cloud Scientific Workflows

Author

Gerlach, Wolfgang ; Wei Tang ; Keegan, Kevin ; Harrison, Travis ; Wilke, Andreas ; Bischof, Jared ; D´Souza, Meenakshi ; Devoid, Scott ; Murphy-Olson, Daniel ; Desai, Narayan ; Meyer, Folker

Author_Institution

Univ. of Chicago, Chicago, IL, USA

fYear

2014

fDate

21-21 Nov. 2014

Firstpage

25

Lastpage

32

Abstract

Recently, Linux container technology has been gaining attention as it promises to transform the way software is developed and deployed. The portability and ease of deployment makes Linux containers an ideal technology to be used in scientific workflow platforms. Skyport utilizes Docker Linux containers to solve software deployment problems and resource utilization inefficiencies inherent to all existing scientific workflow platforms. As an extension to AWE/Shock, our data analysis platform that provides scalable workflow execution environments for scientific data in the cloud, Skyport greatly reduces the complexity associated with providing the environment necessary to execute complex workflows.

Keywords

Linux; cloud computing; data analysis; natural sciences computing; virtualisation; AWE/Shock; Docker Linux containers; Skyport; container-based execution environment management; data analysis platform; multicloud scientific workflows; resource utilization inefficiencies; scalable workflow execution environments; scientific data; software deployment problems; Containers; Electric shock; Kernel; Linux; Servers; Virtualization;

fLanguage

English

Publisher

ieee

Conference_Titel

Data-Intensive Computing in the Clouds (DataCloud), 2014 5th International Workshop on

Conference_Location

New Orleans, LA

Type

conf

DOI

10.1109/DataCloud.2014.6

Filename

7017950