Location: Sheraton Centre Toronto (City Hall Room)
| 08:15 - 09:00 | REGISTRATION / CONTINENTAL BREAKFAST
| 09:00 - 09:10 | OPENING REMARKS - Chairperson
| 09:10 - 10:30 | IT'S TIME TO RETHINK YOUR BUFFERPOOL STRATEGY
| Martin Hubel - Martin Hubel Consulting Inc. 10:30 - 10:45
| COFFEE BREAK
| 10:45 - 12:00 | SCALING UP WITH DB2 UDB
| Paul Yu - CIBC 12:00 - 01:15
| LUNCH
| 01:15 - 02:30 | IMPROVING PERFORMANCE WITH DB2/UDB INDEXES
| Leslie Cranston - IBM Toronto Lab. 02:30 - 02:45
| COFFEE BREAK
| 02:45 - 04:00 | BUSINESS INTELLIGENCE SCALAR AGGREGATE FUNCTIONS
| Calisto Zuzarte - IBM Toronto Lab. |
IT'S TIME TO RETHINK YOUR BUFFERPOOL STRATEGY
As applications grow, buffer pools play an increasingly important part in reducing physical I/O, elapsed time and CPU time.
Many organizations have not spent the time to optimize this valuable resource.
Martin will discuss his experiences in tuning buffer pools for large applications including PeopleSoft, production applications, and data warehouses at client sites.
The presentation will include measurements and guidelines, and will cover five main areas:
Martin Hubel, an independent consultant, has worked extensively with DB2 since 1985. Martin develops and teaches advanced DB2 courses and is recognized as a leading authority in the field.
SCALING UP WITH DB2 UDB
This is a case study of Canadian Imperial Bank of Commerce’s experience in scaling up their database. In 1997, CIBC was using SQL Server to support their Java-based application. Operational problems and concern for the database’s ability to scale up led to a switch to DB2 UDB.
The case study covers:
IMPROVING PERFORMANCE WITH DB2/UDB INDEXES
The creation of appropriate indexes is critical to application performance.
However, the decisions that go into creating the 'right' indexes - how many, on what columns, what type - are often non-trivial.
We will discuss in detail the latest index features, and then learn how to understand your database requirements in order to design the 'optimal' indexes for your environment,
as well as how to avoid situations that negatively impact performance.
Leslie is a member of the IBM DBT (Database Technology) Group at the IBM Toronto Lab.
BUSINESS INTELLIGENCE SCALAR AGGREGATE FUNCTIONS
Much work has been done in DB2 UDB to make it easier for applications to handle Business Intelligence (BI) functionality.
In earlier versions, several concepts were introduced into the product.
These included the addition of a number of statistical "scalar" functions and enhancements to the "aggregation" functionality (CUBE, ROLLUP, GROUPING SETS) and the Automatic Summary Table (AST) feature.
This presentation deals with a new concept of "scalar aggregate" BI functions implemented in DB2 V7.1.
These functions allow many of the typical BI functions done in applications to be pushed down closer to the source of the data.
They include ranking functions and sliding window aggregate functions that are very useful for BI applications when analysing and presenting the data.
Calisto is a member of the DBT Optimizer Development Team in the IBM Toronto Lab.