Home | Email Me
About Me | Resume | Research & Publications | Useful Links | Contact Information

Bo Hong's Curriculum Vitae RESEARCH INTERESTS Storage and file systems, including system design, workload modeling, runtime performance modeling and prediction, QoS support, scalability, and dependability. I am also interested in system support for data-intensive computing and high-performance computing, distributed and clustered systems, and performance evaluation. EDUCATION University of California at Santa Cruz, Santa Cruz, CA                                          Ph.D. Computer Engineering, 09/2005 Dissertation title: "Storage and File Systems for MEMS-Based Storage" Advisors: Prof. Darrell D. E. Long and Prof. Scott A. Brandt University of California at Santa Cruz, Santa Cruz, CA M.S. Computer Engineering, 09/2002 Thesis title: "Techniques for Synthetic I/O Workload Generation" Advisor: Prof. Tara M. Madhyastha Purdue University, West Lafayette, IN Graduate study in Statistics, 08/1999 - 05/2000 University of Science and Technology of China (USTC), Hefei, China B.S. Computer Science, 07/1999 Five-year program in the Special Class for the Gifted Young WORK EXPERIENCE   Principal Software Engineer Storage and Server Management Group, Symantec Corporate, Mountain View, CA, 08/2005 - present RESEARCH EXPERIENCE   RESEARCH ASSISTANT Storage Systems Research Center, University of California, Santa Cruz, CA, 06/2002 - 07/2005 Advisors: Prof. Scott A. Brandt and Prof. Darrell D. E. Long MEMS-based storage is an emerging non-volatile secondary storage technology. With fundamentally different architectural designs from magnetic disk, MEMS-based storage promises high performance, low power consumption, high storage volumic density, and several other interesting characteristics. Our research focuses on understanding the impacts of MEMS-based storage on the performance, cost/performance, and reliability of storage systems. Developed an analytical seek time model for MEMS-based storage. Developed Zone-based Shortest Positioning Time First (ZSPTF) request scheduling for MEMS-based storage that achieves SPTF-like performance as well as C-SCAN-like service fairness. Developed two alternative storage architectures that employ MEMS-based storage as another layer in the storage hierarchy, MEMS Write Buffer and MEMS Caching Disk, significantly improving the system performance and cost/performance ratio. Developed the concept of MEMS Storage Enclosures, and evaluated the reliability and economic issues of such enclosures under different maintenance policies. Other research activities include: Developed an analytical model for flash crowd events on the Internet and evaluated the performance of various multi-level caching techniques under these events. Analyzed file system level workloads from large scientific computing applications. RESEARCH INTERN Storage Tank Group, IBM Almaden Research Center, San Jose, CA, 06/2004 - 09/2004 Mentor: Demyn Plantenberg   Providing NFS v3/v4 services through a cluster of NAS gateways is crucial for the wider adoption of the IBM SAN file system, a.k.a. IBM Storage Tank. Investigated load-balancing strategies in the NAS gateway cluster. Possible U.S. patent application to be filed on work in collaboration with Demyn Plantenberg and Dr. Juan C. Gomez. The details of this project are currently under nondisclosure agreement with IBM. RESEARCH INTERN Storage Tank Group, IBM Almaden Research Center, San Jose, CA, 06/2003 - 08/2003 Mentor: Demyn Plantenberg Duplicate data can occupy a substantial portion of a storage system. Addressing unrecognized data redundancy could result in better utilization of storage resources. Our research paper was published in MSST 2004. Measured the degree of data duplication in several on-line file systems for various application environments. Developed a duplicate data block elimination technique for the IBM Storage Tank, based on content-based hashing, copy-on-write, and lazy updates. RESEARCH ASSISTANT Storage Technology Advanced Research, University of California, Santa Cruz, CA, 06/2000 - 06/2002 Advisor: Prof. Tara M. Madhyastha I/O workloads are extremely difficult to characterize and model. We investigated ways to generate synthetic traces that can accurately reproduce disk behaviors with respect to response times. Revealed the irrelevance of long-range dependence in disk traffic. Developed a multifractal-based method that can accurately reproduce local burstiness in I/O arrival patterns. Developed a cluster-based method that can reconstruct high-quality synthetic workloads from a relatively small set of representative intervals from real traces.   UNDERGRADUATE STUDENT PROGRAMMER Artificial Intelligence Group, Dept. of Computer Science, USTC, Hefei, China, 03/1996 - 07/1997 Computer-Aided Instruction (CAI) is a computer-based technology that assists the teaching and learning process. Its cost-effectiveness makes it particularly useful in developing countries where education resources are limited. Led the development of "CAI Software for College Fundamental Physics". Participated in the development of "CAI Simulation Software for College Physics Experiments". TEACHING EXPERIENCE TEACHING ASSISTANT School of Engineering, University of California, Santa Cruz, CA, 09/2000 - 06/2001 Department of Statistics, Purdue University, West Lafayette, IN, 08/1999 - 05/2000 Elementary Statistics PUBLICATIONS Refereed Conference Papers Bo Hong, Scott A. Brandt, Darrell D. E. Long, Ethan L. Miller, and Ying Lin. Using MEMS-Based Storage in Computer Systems -- Device Modeling and Management. In ACM Transactions on Storage (TOS), to appear. Bo Hong, Feng Wang, Scott A. Brandt, Darrell D. E. Long, and Thomas J. E. Schwarz. Using MEMS-Based Storage in Computer Systems -- MEMS Storage Architectures. In ACM Transactions on Storage (TOS), to appear. Bo Hong, Tara M. Madhyastha, and Bing Zhang. Cluster-Based Input/Output Trace Synthesis. In Proceedings of the 24th IEEE International Performance Computing and Communications Conference (IPCCC 2005), Phoenix, AZ, April 2005. A technical report version is available here. Feng Wang, Bo Hong, Scott A. Brandt, and Darrell D. E. Long. Using MEMS-Based Storage to Boost Disk Performance. In Proceedings of the 22nd IEEE / 13th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST 2005), Monterey, CA, April 2005.  Bo Hong and Tara M. Madhyastha. The Relevance of Long-Range Dependence in Disk Traffic and Implications for Trace Synthesis. In Proceedings of the 22nd IEEE / 13th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST 2005), Monterey, CA, April 2005. Bo Hong, Thomas J. E. Schwarz, Scott A. Brandt, and Darrell D. E. Long. Reliability of MEMS-Based Storage Enclosures. In Proceedings of the 12th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS  2004), Volendam, the Netherlands, October 2004, pages 571-579. Bo Hong, Demyn Plantenberg, Darrell D. E. Long, and Miriam Sivan-Zimet. Duplicate Data Elimination in a SAN File System. In Proceedings of the 21st IEEE / 12th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST 2004), College Park, MD, April 2004, pages 301-314. Feng Wang, Qin Xin, Bo Hong, Ethan L. Miller, Darrell D. E. Long, Scott A. Brandt, and Tyce McLarty. File System Workload Analysis For Large Scientific Computing Applications. In Proceedings of the 21st IEEE / 12th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST 2004), College Park, MD, April 2004, pages 139-152. Bo Hong, Scott A. Brandt, Darrell D. E. Long, Ethan L. Miller, Karen A. Glocer, and Zachary N. J. Peterson. Zone-Based Shortest Positioning Time First Scheduling for MEMS-Based Storage Devices. In Proceedings of the 11th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS 2003), Orlando, FL, October 2003, pages 104-113. Ismail Ari, Bo Hong, Ethan L. Miller, Scott A. Brandt, and Darrell D. E. Long. Managing Flash Crowds On the Internet. In Proceedings of the 11th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS 2003), Orlando, FL, October 2003, pages 246-249. A more interesting technical report version is available here.  Technical Reports Ismail Ari, Bo Hong, Ethan L. Miller, Scott A. Brandt, and Darrell D. E. Long. Modeling, Analysis and Simulation of Flash Crowds on the Internet. Technical Report UCSC-CRL-03-15, Storage Systems Research Center, University of California at Santa Cruz, February 2004. Bo Hong and Scott A. Brandt. An Analytical MEMS Seek Time Model. Technical Report UCSC-CRL-02-31, Storage Systems Research Center, University of California at Santa Cruz, September 2002. Bo Hong. Techniques for Synthetic I/O Workload Generation. Master Thesis. Technical Report UCSC-CRL-02-16, University of California at Santa Cruz, March 2002. SOFTWARE RELEASED CAI Software for College Fundamental Physics. Chinese Higher Education Publishing House, 1997. CAI Simulation Software for College Physics Experiments. Chinese Higher Education Publishing House, 1996. SERVICE Referee for Supercomputing, SIGMetrics, FAST, ICDCS, MSST, MASCOTS, PACT, and others. Computer Science Faculty Recruiting Committee, 2004. HONORS Undergraduate fellowship in USTC, 1995-1999. REFERENCES Available upon request.