yun tian and xiao qin
play

Yun Tian and Xiao Qin Computer Science and Software Engineering - PowerPoint PPT Presentation

Yun Tian and Xiao Qin Computer Science and Software Engineering Department Auburn University Auburn, AL 36849 Large-scale data processing Scalability, availability, performance, security 2 8/22/2011 Security is achieved at the


  1. Yun Tian and Xiao Qin Computer Science and Software Engineering Department Auburn University Auburn, AL 36849

  2.  Large-scale data processing  Scalability, availability, performance, security 2 8/22/2011

  3. Security is achieved at the cost of performance degradation P S P Security S Overheads 3 8/22/2011

  4.  (m, n) Secret sharing - e.g., (2, 3)  File Fragmentation Motivation: Storage nodes in a distributed system have heterogeneous vulnerabilities. 4 8/22/2011

  5.  A wide variety of: ◦ Hardware (e.g., SSDs, HDDs, Tapes) ◦ Software (e.g, HDFS, Lustre, PVFS)  Heterogeneities affect performance Can we leverage heterogeneity features to improve security for distributed systems? 5 8/22/2011

  6. File F has 3 fragments: F a a, b, c Fragment a Fragment b Fragment c F(i)= j mod n 6 8/22/2011

  7. Problem? Reconstruct Fragment a Fragment b Fragment c File F  7 8/22/2011

  8.  The Architecture of a Cluster Storage subsystems Head (or Storage Area Network) Internet Node Client Network switch Computing nodes 8 8/22/2011 8/22/2011 8

  9. 9 8/22/2011

  10. Reconstruct File F R R1 Rk f1 fk 10 8/22/2011

  11. A file’s fragment -allocation decisions are guided by four policies • Multiple server groups • Store fragments of a file across as many different server groups as possible • Integrate the (m n) secret sharing scheme • Allocate file fragments within a sub-system to improve I/O performance 11 8/22/2011

  12.  Diversity make sense? • A team with diversity make creativity; • A system with diversity may improve security; • We divide storage nodes of a system into different “server type” based on their different security level or strategy caused by hardware or software; 12 8/22/2011

  13.  Store fragments of a file across as many different server-type groups as possible 13 8/22/2011

  14.  Allocate file fragments to improve I/O performance • Allocating fragments of a file into different storage clusters can degrade performance. • Our S-FAS scheme attempts to allocate fragments to storage nodes within a cluster. 14 8/22/2011

  15. Fragmentation technique To improve Heterogeneous Natures the assurance level Secret sharing scheme(m n) 15 8/22/2011

  16. 16 8/22/2011

  17. Static Dynamic 17 8/22/2011

  18. 18 8/22/2011

  19. We gradually increase system size from 45 to 70 by increments of 5, keep k at 3,and also vary m from 4 to 8. 19 8/22/2011

  20. 20 8/22/2011

  21. 21 8/22/2011

  22. 22 8/22/2011

  23. 23 8/22/2011

  24.  Heterogeneous vulnerability: storage nodes are classified into different server groups based upon their vulnerabilities  S-FAS : a secure fragmentation allocation scheme  Storage assurance and dynamic assurance models  A prototype in which S-FAS was implemented 24 8/22/2011

  25.  Consider data replications to enhance reliability and performance  Authorization/Authentication mechanisms in the prototype.  Encryption/Decryption Mechanisms in the prototype. 25 8/22/2011

  26. Google: slideshare Xiao Qin ‹#›

  27. Other Research Presentations and Projects http://www.eng.auburn.edu/~xqin

  28. My webpage http://www.eng.auburn.edu/~xqin

  29. Download Slides at slideshare http://www.slideshare.net/xqin74

  30. Thank you! 30 8/22/2011

  31. Suggestions & Questions? 31 8/22/2011

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend