|
CornellUniversity Ithaca, NY 14853 |
Phone 607-255-9537 Fax 607-255-4428 E-mail mayr@cs.cornell.edu |
Tobias Mayr
|
Objective |
Position as a researcher, starting in August 2001. |
|
|
Academic Background |
1996 - 2001 Cornell University Ithaca, NY, USA PhD Studies in Computer Science Thesis on adaptive parallel data processing, Advisors
Prof. Praveen Seshadri and My thesis work evaluates new distributed execution techniques in query processing systems (see research statement). The systems component of my work was within the context of the Cornell Predator OR-DBMS. I integrated the Java VM and client-site processing into the execution engine and rebuilt the Predator system as a parallel execution environment. I was teaching assistant for three courses at Cornell: an introductory class, a class on algorithms, and one on discrete structures. I gave guest lectures in advanced database courses and seminars. 1993 - 1996 Technische
Universität München Germany Vordiplom
in Computer Science, Diploma Thesis on Mobile Network Algebras, Advisors
Dr.Radu Grosu, Prof. Manfred Broy, Minor in Computational Linguistics and
Philosophy of Language. I
worked on software specification parsing and object-oriented type-checkers. |
|
|
Internships |
Fall 2000 Microsoft BARC San Francisco, CA, USA I worked with Dr.
Jim Gray to
design Summer 1999 IBM Research San Jose, CA, USA I worked with Dr. Mary Tork Roth on the
design and implementation of OLEDB data source integration for the Data
Joiner distributed
database product. |
|
|
References |
References are available upon request from: Prof. Praveen Seshadri, Cornell University Prof. Johannes Gehrke, Cornell University Dr. Jim Gray, Microsoft Corporation |
|
|
Research |
My research interest is
data processing in asymmetric parallel and distributed environments. Future computer systems
will process vast amounts of complex data in parallel environments that
consist of many heterogeneous components. Active storage and network
components, clients and external sites contribute their data, functionality
and processing power to make data processing more scalable, flexible and
powerful. These highly non-uniform processing environments must be integrated
to form reliable and scalable data processing systems. My PhD work shows how
new processing techniques can be used to allow parallel processing of data
while adapting to the heterogeneity of the processing components. I
parallelized the execution engine of an existing object-relational database
system as a prototype for experimentation with the new parallel query
processing techniques. My work demonstrates processing on asymmetric nodes in
a cluster as well as the integration of external functionality on server and
client sites. The underlying idea is that parallel systems must be aware of
the heterogeneity of their components to process data efficiently and to
fully integrate all components. In my future work, I
want to apply these ideas to new environments, like mobile device networks,
systems with intermittently connected clients or peer-to-peer architectures.
I want to apply my ideas in actual applications to develop them as realistic
solutions to real-world problems. |
|
Publications |
Tobias Mayr and
Praveen Seshadri. Client-Site Query
Extensions. SIGMOD 1999, pages 347-358 Greg Czajkowski, Tobias
Mayr, Praveen Seshadri, and Thorsten von Eicken. Resource Control for Database Extensions. COOTS 1999. Mike Godfrey, Tobias
Mayr, Praveen Seshadri, and Thorsten von Eicken. Secure and Portable Database Extensibility. SIGMOD 1998, pages
390-401 Tobias Mayr: A Model for Mobile Networks with
Synchronous and Asynchronous Operators. Diplomarbeit, Technische
Universitaet Muenchen, 1996. |