Attack method for big data processing platform Hadoop

A technology of big data processing and hadoop clusters, applied in electrical components, transmission systems, etc., can solve the problems of no security vulnerability analysis, no specific implementation plan for attacks, etc., and achieve the effect of low cost

Inactive Publication Date: 2020-05-15
江苏海平面数据科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The paper "Fu X , Gao Y , Luo B , et al. Security Threats to Hadoop: DataLeakage Attacks and Investigation[J]. IEEE Network, 2017, 31(2):67-71." Although it has detailed data leakage attacks analysis, but did not give a specific implementation of the attack, and did not compare it with other types of attack schemes
Paper "Wang J , Wang T , Yang Z , et al. SEINA: AStealthy and Effective Internal Attack in Hadoop Systems[C] / / International Conference on Computing, Networking and Communications (ICNC 2017). IEEE,2017: 525-530." Proposed an attack scheme inside Hadoop, but it also did not analyze other security vulnerabilities, and compared it with other attack schemes

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Attack method for big data processing platform Hadoop
  • Attack method for big data processing platform Hadoop
  • Attack method for big data processing platform Hadoop

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0063] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0064] The present invention assumes that the cluster has deployed some defense mechanisms, and the proposed attack will follow the security policy set by the cluster. In this invention, the Hadoop data processing framework is used as the target platform of the attack, but the attack can also be applied to other similar frameworks.

[0065] The first attack model, MR speculative task scheduling attack (MR Speculative Attack), is a scheduling algorithm for mapreduce speculative tasks [4]; it uses the loopholes in hadoop scheduling method to extend the execution time of jobs.

[0066] The second attack model, HeartBeat Attack, is aimed at the communication mechanism between the master node and the slave node; the scheduling time of map and reduce tasks is increased by extending the heartbeat time.

[0067] The third attack model data attack (Dat...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention belongs to the field of big data security in the field of computers, and relates to an attack method for a big data processing platform Hadoop. The attack method is provided with a speculative task scheduling attack model, a heartbeat attack model, a data attack model and a Shuffle attack model, wherein the speculative task scheduling attack model can slow down the execution of individual tasks for speculative task scheduling algorithm attacks in a cluster, and then slow down the execution of the whole operation; the heartbeat attack model prolongs heartbeat sending time to timeout time for attacks of a heartbeat mechanism in the cluster so as to influence performance of the cluster; the data attack model attacks intermediate data in the mapreduce task execution process, andcarries out operations such as modification, deletion and addition on the content of an intermediate result so as to influence the whole job execution result; and the Shuffle attack model attacks thecluster at a Shuffle stage in the task execution process, and prolongs the execution time of the whole task by prolonging the execution time of the Shuffle stage.

Description

technical field [0001] The invention belongs to the field of big data security in the computer field, and relates to an attack method for Hadoop, a big data processing platform. Background technique [0002] With the rise and development of cloud computing and big data, more and more enterprises and users use big data processing platforms such as Hadoop and Spark to process and analyze massive data. For example, platform-as-a-service (PAAS) platforms provide application development and runtime environments through cloud computing technology. Since the results of big data analysis provide critical information for a large number of applications, the security of data processing frameworks becomes a very important issue. [0003] Existing big data processing platforms are distributed platforms. All operations are not executed on one machine, but are distributed on different machines, which can shorten the execution time of jobs and improve execution efficiency. However, due t...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): H04L29/06
CPCH04L63/1441H04L63/20
Inventor 孙杰沈祥红王亚炜
Owner 江苏海平面数据科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products