Browse Prior Art Database

Query Optimizations Surrounding Different Memory Types

IP.com Disclosure Number: IPCOM000215272D
Publication Date: 2012-Feb-23
Document File: 1 page(s) / 20K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed are query optimizations surrounding different memory types.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 56% of the total text.

Page 01 of 1

Query Optimizations Surrounding Different Memory Types

Traditional database computing has existed now for almost 30 years and database optimization surrounding query performance has been on the forefront of everyone's mind for all 30 years. The common theme of optimizers is to find the fastest access plan a query can use and then run the query using that access plan. Optimizers take into consideration the amount of I/O, CPU needs and how much memory is required to perform such as function. Memory considerations and how much memory is utilized by a given query are key decision points to any optimizer.

    What optimizers do not do well is take into consideration that complex database operations such as JOINS and the like can only run as fast as the slowest part. Meaning if a query is broken up into multiple parts, the query cannot complete until all parts are processed. Therefore, speeding up certain parts of a "distributed" query is advantageous, whereby speeding up the already fast parts provides little to no value to the overall execution time.

    With the advent of new technology where all memory is not created equal; i.e., flash memory versus traditional memory types. Proposed here is taking database optimization to the next level. Foreseen is an environment where different parts of a query will execute with different types of memory such that a query's overall run-time performance will not degrade.

    This invention, therefore, is a method by which a query optimizer...