PhD Thesis Proposal Defence


"Cache-Oblivious Query Processing"

by

Mr. Bingsheng HE


Abstract:

Cache-oblivious techniques, proposed in the theory community, are
aware of the existence of a multi-level memory hierarchy but do not
assume any knowledge about the parameter values of the hierarchy,
such as the number of levels in the hierarchy, the capacity and the
block size of each level. These cache oblivious techniques are
proved to have optimal bounds on the amount of data transferred
between any two adjacent levels of an arbitrary memory hierarchy.
Moreover, this optimal performance is achieved without any hardware
platform specific tuning.  These properties are highly attractive to
autonomous databases, especially because the hardware architectures
are becoming increasingly complex and diverse.

We propose to develop EaseDB, the first cache-oblivious query
processor for main memory relational databases. Since traditional
database systems are intrinsically aware of cache parameters, e.g.,
the memory page size, our cache-oblivious approach requires us to
rethink the query processing architecture and implementation.  In
this proposal, we present the architectural design of EaseDB, report
our initial results and outline our future research plan.


Date:     		Tuesday, 17 April 2007

Time:                   10:00a.m.-12:00noon

Venue:                  Room 4480
			lifts 25-26

Committee Members:      Dr. Qiong Luo (Supervisor)
                        Prof. Lionel Ni (Chairperson)
			Prof. Dik Lun Lee
			Prof. Frederick Lochovsky


**** ALL are Welcome ****