The objective of this document is to guide anyone new to the implementation of workload management within the DB2 for Linux, UNIX, and Windows product through the step-by-step process needed to establish an initial configuration that is designed to help ensure the stability and predictability of the database system as a whole.

The following are the two main lessons learned from experience with regards to workload management:

1. Having a workload management plan or configuration in place for your system provides the following significant advantages:

   -A database system cannot become swamped and unresponsive due to low priority, complex work consuming too much resource

   -The root cause of many issues in DB2 environments are resolved with the implementation of workload management principles

2. A workload management configuration that is too complex makes a system difficult to monitor and manage
   
   -The "Keep It Simple, Silly" (KISS) principle is still the best approach, and complexity should be added only when needed

This document presents a set of definitions representing the different stages of maturity for a workload management configuration in a DB2 for Linux, UNIX, and Windows database. These stages range from stage 0 through to the advanced stage 3 configuration. A specific configuration template and process is provided as part of these best practices to enable customers to progress from a stage 0 configuration to a stage 2 configuration. General descriptions and advice are also given about common stage 3 scenarios.

It is recommended that all DB2 customers, with a data warehouse, implement at least a stage 2 workload management configuration and this goal is the primary focus of this document. A stage 2 configuration focuses on stabilizing the overall system behavior. Some customers might require a more advanced, stage 3 configuration to address their specific application performance objectives or to support their business and IT philosophies regarding how users are serviced.

The focus of this document is to describe how best to implement a successful workload management solution using DB2 for Linux, UNIX, and Windows, Version 9.7.4 or higher. The contents of this document reflect the latest experiences of IBM® field personnel and customers within the data warehouse arena from which the vast majority of reported feedback for DB2 workload management has been received.

Using a staged approach, this document guides you through the steps needed to implement the best practices workload management configuration on DB2 for Linux, UNIX, and Windows with sufficient controls to help ensure a stable, predictable system for most data warehouse environments. This initial configuration is intended to be a good base upon which additional tuning and configuration changes can be implemented, as needed, for you to achieve your specific workload management objectives.

The document assumes a novice beginner and describes the individual steps and mechanisms at each point. A more experienced user can condense many of the listed steps to move from stage 1 to stage 2, making the transition in days of elapsed time rather than weeks as the suggested timeline indicates in a later section.

Although you can use SQL DDL statements to directly interact with the DB2 database manager, the IBM OptimTM Performance Manager 4.1.1 product provides a simplified interface through its detailed configuration editor which can be used to make the interactions less onerous.

The steps outlined in this document are focused on the efficiency of the system as a whole, regardless of where the work itself comes from. It is important to note that achieving the goal of a stable system might not necessarily also result in the achievement of any individual application service-level agreement (SLA) or specific performance objectives for queries. These more granular objectives might require subsequent changes to the workload management configuration, such as outlined in the section on stage 3 scenarios, which is outside the main scope of this document.

This paper is not a tutorial on DB2 workload management capabilities and does not attempt to provide comprehensive guidance in addressing all possible scenarios where DB2 workload management might be employed. It also does not cover all features within the DB2 product that might be of use in controlling resource consumption. The scope of this paper is focused on describing the system stabilization approach in some detail and provides some general guidance for common advanced scenarios.