Columbia Query Optimizer

The Columbia Query Optimization Project is a joint research project of David Maier from the Oregon Graduate Institute and Leonard Shapiro from Portland State University . It is part of the "Query Optimization Engineering" project funded by the National Science Foundation, to Portland State University (NSF # IRI-9610013) and to Oregon Graduate Institute (NSF #IRI-9619977). This site introduces the Columbia project and provides access to its documentation and source code.

Introduction to Query Optimization
(Chapter 13 of [Ram97])

       When a query is presented to a database system, the system should have a mechanism to produce a good plan for that query using the resources available, e.g., the catalogs,indexes etc. By good plan we mean that the plan produced thereby is correct and has low cost according to the cost model. This is important because the cost difference between a good versus a bad plan can be substantial. This is where the use of query optimizers gains significance. Even though research on optimizers has been going on for more than a decade, making an extensible and efficient optimizer is still a complex task.

Reading material on Columbia and related research

       In order to understand Columbia, one should look at the following documents in the order that they are listed.
Obtaining the source code
Using the Columbia Optimizer
beginning screen(Fig 1.1)
project settings
project settings(Fig 1.2)
set Active configuration
-uration(Fig 1.3)
building and executing
building and executing(Fig 1.4)
Options Window
Option Window(Fig 2.1)
trace output
output(Fig 2.2)
viweing the output
output(Fig 2.3)

Frequently Asked Questions

Contact: Kavita