Yang Wang
Office: ACES 6SEo4D
Department of Computer Science
University of Texas at Austin
Email: yangwang AT cs dot utexas dot edu
I'm a fourth year Ph.D. student in the LASR group of the department of computer science. My advisor is Prof. Mike Dahlin. My current interests are in distributed systems and fault tolerance.

I got my bachelor's degree (July 2005) and master's degree (Jan. 2008) in Computer Science from Tsinghua University.

  • Curriculum Vitae [PDF]

    Research
  • Salus: Robust and Scalable Block Store. The target is to enhance the robustness of those highly scalable storage systems like Bigtable/HBase. First, Salus performs end-to-end verification and incorporates a pipelined commit protocol to achieve this without hurting performance. Second, Salus replicates tablet server/region server to eliminate single point of failure and reduce network bandwidth consumption. Finally, Salus use lazy recovery in erasure coding to reduce its recovery cost.
  • Gnothi: Efficient and Available Storage Replication. The goals of this project are 1) efficiency: write to f+1 nodes and read from 1, which are both minimal if the system can tolerate f faults. 2) availability: can use short timeout without breaking consistency guarantee, like in asynchronous replication. Gnothi achieves both by separating data and metadata: replicating data to f+1 nodes for efficiency and replicating metadata to 2f+1 nodes through a Paxos-like protocol for availability.
  • Eve: Execute-Verify Replication for Multi-Core Servers. The key challenge of replicating a multithreaded application is its nondeterministic behavior. Eve first guesses which requests have no conflicts and thus can be executed in parallel. Then it speculates the results of the execution and if they diverge, Eve rolls back all replicas and re-execute sequentially.
  • UpRight Cluster Services. The UpRight project builds a library and framework for building distributed systems with high availablity. The library can be configured to tolerate both crash failures and Byzantine failures. We've modified the Hadoop Distributed File System (HDFS) and Zookeeper coordination and lock service to use the UpRight library.
    • Publications
  • "Robustness in the Salus scalable block store". Y. Wang, M. Kapritsos, Z. Ren, P. Mahajan, J. Kirubanandam, L. Alvisi, and M. Dahlin, to appear in NSDI 2013.
  • "All about Eve: Execute-Verify Replication for Multi-Core Servers". M. Kapritsos, Y. Wang, V. Quema, A. Clement, L. Alvisi, and M. Dahlin, in 2012 USENIX Symposium on Operating Systems Design and Implementation (OSDI12), Hollywood, CA, October 2012.
  • "Gnothi: Separating Data and Metadata for Efficient and Available Storage Replication". Y. Wang, L. Alvisi, and M. Dahlin, in 2012 USENIX Annual Technical Conferecne, Boston, MA, June 2012.
  • "SOPA: Selecting the Optimal Policy Adaptively". Yang Wang, Jiwu Shu, Guangyan Zhang, Wei Xue, Weimin Zheng, ACM transactions on storage, Volume 6 Issue 2, July 2010.
  • "UpRight Cluster Services". A. Clement, M. Kapritsos, S. Lee, Y. Wang, L. Alvisi, M. Dahlin, T. Riche, Proceedings of the 22 nd ACM Symposium on Operating Systems Principles (SOSP), Oct 2009 .
    • Teaching
    2008 Fall, CS352 Computer Systems Architecture by Prof. Herbert Schwetman
    2009 Spring, CS345 Programming Languages by Prof. Philip Cannata
    Others
    I like basketball, ping-pong, Weiqi (Go), and travelling.