Scientific workflow execution system based on mimic defense in the cloud environment

With more large-scale scientific computing tasks being delivered to cloud computing platforms, cloud workflow systems are designed for managing and arranging these complicated tasks. However, multi-tenant coexistence service mode of cloud computing brings serious security risks, which will threaten...

Full description

Saved in:
Bibliographic Details
Published in:Frontiers of information technology & electronic engineering 2018-12, Vol.19 (12), p.1522-1536
Main Authors: Wang, Ya-wen, Wu, Jiang-xing, Guo, Yun-fei, Hu, Hong-chao, Liu, Wen-yan, Cheng, Guo-zhen
Format: Article
Language:English
Subjects:
Cloud computing
Communications Engineering
Computer Hardware
Computer Science
Computer Systems Organization and Communication Networks
Cybersecurity
Electrical Engineering
Electronics and Microelectronics
Heterogeneity
Instrumentation
Networks
Redundancy
Task scheduling
Workflow management systems
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With more large-scale scientific computing tasks being delivered to cloud computing platforms, cloud workflow systems are designed for managing and arranging these complicated tasks. However, multi-tenant coexistence service mode of cloud computing brings serious security risks, which will threaten the normal execution of cloud workflows. To strengthen the security of cloud workflows, a mimic cloud computing task execution system for scientific workflows is proposed. The idea of mimic defense contains mainly three aspects: heterogeneity, redundancy, and dynamics. For heterogeneity, the diversities of physical servers, hypervisors, and operating systems are integrated to build a robust system framework. For redundancy, each sub-task of the workflow will be executed simultaneously by multiple executors. Considering efficiency and security, a delayed decision mechanism is proposed to check the results of task execution. For dynamics, a dynamic task scheduling mechanism is devised for switching workflow execution environment and shortening the life cycle of executors, which can confuse the adversaries and purify task executors. Experimental results show that the proposed system can effectively strengthen the security of cloud workflow execution.
ISSN:2095-9184
2095-9230
DOI:10.1631/FITEE.1800621