Browsing by Author "Xiaosong Ma, Committee Member"

Now showing 1 - 7 of 7

Efficient Subsequence Matching with LCS
(2007-10-08) Han, Tae Sik; Jaewoo Kang, Committee Co-Chair; Rada Chirkova, Committee Co-Chair; Xiaosong Ma, Committee Member; Ting Yu, Committee Member
Grid Service Data Needed for Estimation of Reliability in Scientific Workflow Systems
(2004-06-11) Colonnese, Daniel; Mladen Vouk, Committee Chair; Xiaosong Ma, Committee Member; Ting Yu, Committee Member; Gregor von Laszewski, Committee Member
The emerging technologies of grid computing, web services, and service-oriented workflows will enable scientific projects to be conducted on a larger scale than ever before. Scientific workflows are often very dynamic in both structure and persistency, complex, and may involve a lot of interaction, very large data flows, and may change on very short notice. Given modern information technology tools, they can be constructed by combining dispersed network accessible services into virtual organizations. Within a scientific workflow environment, metadata or grid service data is necessary for service consumer application to discover services and for services to publish their properties. Software reliability engineering within service oriented workflow systems is still a largely unexplored area. Emerging standards such as WS-Reliability and WS-Agreement are not yet practical to function in grid service environment. Therefore, methods of leveraging reliability service data with the existing aggregation mechanisms in Globus Toolkit 3, need explored further exploration. The Open Grid Service Architecture (OGSA) allows registries of services. Service data information in the registries can be used for soft-state management, keeping track of metadata for service instances created from application factories, and so on. A service aggregation registry subscribes to services produced from a number of factories. Service data is express through a proposed XSD namespace shared vocabulary. Discovery policy is expressed in XPath queries. This thesis proposes one of the first architectures and implementations of a grid service registry information specifically intended for support of fault tolerance and service replica selection in scientific workflow application domain. The registry uses the Globus Toolkit 3 and is made available as an OGSI compliant grid service.
Innovative Schemes to Suppport Incremental Checkpointing.
(2010-08-09) Vasavada, Manav; Rainer Mueller, Committee Chair; Xiaosong Ma, Committee Member; Xuxian Jiang, Committee Member
Memory Trace Compression and Replay for SPMD Systems using Extended PRSDs.
(2010-08-05) Budanur Ramanna, Sandeep; Rainer Mueller, Committee Chair; Xiaosong Ma, Committee Member; Xuxian Jiang, Committee Member
Preserving Time in Large-Scale Communication Traces
(2008-08-19) Ratn, Prasun; Dr Frank Mueller, Committee Chair; Xiaohui Gu, Committee Member; Xiaosong Ma, Committee Member
Analyzing the performance of large-scale scientific applications is becoming increasingly difficult due to the sheer size of performance data gathered. Recent work on scalable communication tracing applies online interprocess compression to address this problem. Yet, analysis of communication traces requires knowledge about time progression that cannot trivially be encoded in a scalable manner during compression. We develop scalable time stamp encoding schemes for communication traces. At the same time, our work contributes novel insights into the scalable representation of time stamped data. We show that our representations capture sufficient information to enable what-if explorations of architectural variations and analysis for path-based timing irregularities while not requiring excessive disk space. We evaluate the ability of several time-stamped compressed MPI trace approaches to enable accurate timed replay of communication events. We evaluate our timing methods against various stages of compression to study the effects of compression on timing accuracy. Specifically, we measure accuracy by comparing the original application execution times with the compressed trace replay times. Our lossless traces are orders of magnitude smaller, if not near constant size, regardless of the number of nodes while preserving timing information suitable for application tuning or assessing requirements of future procurements. Our results prove time-preserving tracing without loss of communication information can scale in the number of nodes and time steps, which is a result without precedent.
Query Size Estimation through Sampling
(2005-01-06) Kim, Kyoung-Hwa; Jaewoo Kang, Committee Member; Xiaosong Ma, Committee Member; Rada Y. Chirkova, Committee Chair
Current Database management systems (DBMS) handle huge amounts of data and need fast query response time. DBMSs apply several strategies to execute user queries. Query optimizers in DBMSs compare costs for these strategies and choose the cheapest one. Materialized views are suggested to enhance query response time as one of the strategies in DBMSs. Cost of each strategy has to be accurately estimated to choose right strategy. Because the materialized view is stored as a one table, we consider that sequential scan is used for executing of the materialized view. Therefore, I/O and CPU costs to execute materialized views depend on the number of tuples for the result. Hence, Our focus will be the accuracy of estimation of the number of tuples in materialized views; that is a query-size estimation. Many researches have been proposed to find methods to estimate the cost of query. This thesis reviews these researches and compares good and bad aspects for each cost estimation method. We choose size estimation methods that are more accurate than others to implement. We suggest various query environments for experiments. We suggest a guideline by experimental results.
A Video Coding Approach to Compressive Distributed System Monitoring.
(2010-08-27) Venkatesh, Vinay; Xiaohui Gu, Committee Chair; Xiaosong Ma, Committee Member; Rudra Dutta, Committee Member