This article is one in a series that focuses on improving performance in Db2 Warehouse. The first part provides some background information and lays out the first and most important aspect of query optimization, namely, cardinality estimation or estimating the number of rows in an execution plan.
I spend every day of my life since 1990 as a Db2 DBA both on Db2 for z/OS and Db2 for LUW. During that time, I’ve collected a lot of knowledge that makes my job easier and gets easier and gets the answers to the developer’s questions in a timely fashion.
As many other I got challenged to set up a database architecture. Not only database objects related, describing the different environments was also asked and there was an urgent need for a decent security setup to support rollouts in general.
You can create profiles to selectively control groups of distributed (DBAT) threads and connections, set thresholds for query acceleration (not covered in this article) or provide override properties to model your test subsystem to reflect the computing power of your production environment.