SEINA: A stealthy and effective internal attack in Hadoop systems

Jiayin Wang, Teng Wang, Zhengyu Yang, Ying Mao, Ningfang Mi, Bo Sheng

Research output: Chapter in Book/Report/Conference proceedingConference contribution

16 Citations (Scopus)

Abstract

Big data processing frameworks such as Hadoop [1] are now widely adopted, however the security issues in large scale systems have not been well studied yet. Unlike prior work on data privacy and protection, this paper investigates a potential attack from a compromised internal node against the overall system performance. We develop an effective attack launched from the compromised node that can significantly degrade the data processing performance of the cluster without being detected and blacklisted for job execution, also present a mitigation scheme that protects a Hadoop system from such attack. The results of experiments show that this attack greatly slows down the job executions in the native Hadoop system even with some basic defense mechanisms, however, our mitigation schem can keep the whole cluster running efficiently under such attack.

Original languageEnglish
Title of host publication2017 International Conference on Computing, Networking and Communications, ICNC 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages525-530
Number of pages6
ISBN (Electronic)9781509045884
DOIs
StatePublished - 10 Mar 2017
Event2017 International Conference on Computing, Networking and Communications, ICNC 2017 - Silicon Valley, United States
Duration: 26 Jan 201729 Jan 2017

Publication series

Name2017 International Conference on Computing, Networking and Communications, ICNC 2017

Other

Other2017 International Conference on Computing, Networking and Communications, ICNC 2017
CountryUnited States
CitySilicon Valley
Period26/01/1729/01/17

Fingerprint

Data privacy
Large scale systems
Experiments
Big data

Cite this

Wang, J., Wang, T., Yang, Z., Mao, Y., Mi, N., & Sheng, B. (2017). SEINA: A stealthy and effective internal attack in Hadoop systems. In 2017 International Conference on Computing, Networking and Communications, ICNC 2017 (pp. 525-530). [7876183] (2017 International Conference on Computing, Networking and Communications, ICNC 2017). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCNC.2017.7876183
Wang, Jiayin ; Wang, Teng ; Yang, Zhengyu ; Mao, Ying ; Mi, Ningfang ; Sheng, Bo. / SEINA : A stealthy and effective internal attack in Hadoop systems. 2017 International Conference on Computing, Networking and Communications, ICNC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 525-530 (2017 International Conference on Computing, Networking and Communications, ICNC 2017).
@inproceedings{204b61b70fff40d59ecffdfc1348ca33,
title = "SEINA: A stealthy and effective internal attack in Hadoop systems",
abstract = "Big data processing frameworks such as Hadoop [1] are now widely adopted, however the security issues in large scale systems have not been well studied yet. Unlike prior work on data privacy and protection, this paper investigates a potential attack from a compromised internal node against the overall system performance. We develop an effective attack launched from the compromised node that can significantly degrade the data processing performance of the cluster without being detected and blacklisted for job execution, also present a mitigation scheme that protects a Hadoop system from such attack. The results of experiments show that this attack greatly slows down the job executions in the native Hadoop system even with some basic defense mechanisms, however, our mitigation schem can keep the whole cluster running efficiently under such attack.",
author = "Jiayin Wang and Teng Wang and Zhengyu Yang and Ying Mao and Ningfang Mi and Bo Sheng",
year = "2017",
month = "3",
day = "10",
doi = "10.1109/ICCNC.2017.7876183",
language = "English",
series = "2017 International Conference on Computing, Networking and Communications, ICNC 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "525--530",
booktitle = "2017 International Conference on Computing, Networking and Communications, ICNC 2017",

}

Wang, J, Wang, T, Yang, Z, Mao, Y, Mi, N & Sheng, B 2017, SEINA: A stealthy and effective internal attack in Hadoop systems. in 2017 International Conference on Computing, Networking and Communications, ICNC 2017., 7876183, 2017 International Conference on Computing, Networking and Communications, ICNC 2017, Institute of Electrical and Electronics Engineers Inc., pp. 525-530, 2017 International Conference on Computing, Networking and Communications, ICNC 2017, Silicon Valley, United States, 26/01/17. https://doi.org/10.1109/ICCNC.2017.7876183

SEINA : A stealthy and effective internal attack in Hadoop systems. / Wang, Jiayin; Wang, Teng; Yang, Zhengyu; Mao, Ying; Mi, Ningfang; Sheng, Bo.

2017 International Conference on Computing, Networking and Communications, ICNC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 525-530 7876183 (2017 International Conference on Computing, Networking and Communications, ICNC 2017).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - SEINA

T2 - A stealthy and effective internal attack in Hadoop systems

AU - Wang, Jiayin

AU - Wang, Teng

AU - Yang, Zhengyu

AU - Mao, Ying

AU - Mi, Ningfang

AU - Sheng, Bo

PY - 2017/3/10

Y1 - 2017/3/10

N2 - Big data processing frameworks such as Hadoop [1] are now widely adopted, however the security issues in large scale systems have not been well studied yet. Unlike prior work on data privacy and protection, this paper investigates a potential attack from a compromised internal node against the overall system performance. We develop an effective attack launched from the compromised node that can significantly degrade the data processing performance of the cluster without being detected and blacklisted for job execution, also present a mitigation scheme that protects a Hadoop system from such attack. The results of experiments show that this attack greatly slows down the job executions in the native Hadoop system even with some basic defense mechanisms, however, our mitigation schem can keep the whole cluster running efficiently under such attack.

AB - Big data processing frameworks such as Hadoop [1] are now widely adopted, however the security issues in large scale systems have not been well studied yet. Unlike prior work on data privacy and protection, this paper investigates a potential attack from a compromised internal node against the overall system performance. We develop an effective attack launched from the compromised node that can significantly degrade the data processing performance of the cluster without being detected and blacklisted for job execution, also present a mitigation scheme that protects a Hadoop system from such attack. The results of experiments show that this attack greatly slows down the job executions in the native Hadoop system even with some basic defense mechanisms, however, our mitigation schem can keep the whole cluster running efficiently under such attack.

UR - http://www.scopus.com/inward/record.url?scp=85017344891&partnerID=8YFLogxK

U2 - 10.1109/ICCNC.2017.7876183

DO - 10.1109/ICCNC.2017.7876183

M3 - Conference contribution

AN - SCOPUS:85017344891

T3 - 2017 International Conference on Computing, Networking and Communications, ICNC 2017

SP - 525

EP - 530

BT - 2017 International Conference on Computing, Networking and Communications, ICNC 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Wang J, Wang T, Yang Z, Mao Y, Mi N, Sheng B. SEINA: A stealthy and effective internal attack in Hadoop systems. In 2017 International Conference on Computing, Networking and Communications, ICNC 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 525-530. 7876183. (2017 International Conference on Computing, Networking and Communications, ICNC 2017). https://doi.org/10.1109/ICCNC.2017.7876183