Introduction to NonStop Operations Management

Performance Management
Introduction to NonStop Operations Management–125507
8-7
Step 2—Gathering Performance Information
•
Acceptable performance goals, such as acceptable system response times, deadlines
for batch jobs, and transaction volumes.
•
Acceptable level of service when the system is stressed, such as during peak periods
and during partial equipment failures.
•
The priorities for resource allocation.
Once the requirements are specified, application developers can instrument applications
so that the necessary measurements are available, and the operations staff can proceed
with measuring performance.
Step 2—Gathering Performance Information
Before a system’s performance can be analyzed and optimized, the performance-
analysis-and-tuning staff must gather relevant performance information. This involves:
•
Collecting information about system and application configurations.
•
Deciding what to measure.
•
Deciding when to take measurements. Typically, the staff takes measurements
throughout each day in order to understand how changes in workload affect the
performance of the systems.
•
Deciding how to take measurements. Tandem provides a number of tools to help
measure performance. The tools can help the staff:
•
Collect statistics on processors, disk drives, and files
•
Measure network traffic
•
Analyze disk space usage
•
Collect information on many other resources
For more information about Tandem tools used for performance analysis and tuning,
refer to “Performance Management Tools,” later in this section.
•
Identifying peak periods.
Step 3—Analyzing Performance Information
Performance analysis involves examining performance information for potential
performance problems and includes the following tasks:
•
Evaluating system and application configurations
•
Creating a workload profile
•
Analyzing resource use
•
Identifying workload imbalances