Frequently Asked Questions

This page has links to various FAQs, listed by topic.

Creating a Service Definition

• How can I create an HTML version of a service definition?
• Why keep number of service classes to a minimum?
• Best treatment of 'loved ones'
• Workloads exploiting enclaves
• Initiators for Unix System Services
• Set goals based on average response time or peak period response time
• Mixing goal types for service class periods with same importance
• Advantages of enterprise-wide policy
• Managing transactions within HSM address space

Managing Online Workloads

• Advantages of response time goals over velocity goals
• Storage protection during periods of low activity
• "Standalone" CICS region where all transactions run in same address space
• Where is CICS CPU usage reported?
• As of CICS 4.1, CPU usage of WLM address space increased noticeably
• With distributed DB2 V4.1, erratic response time in goal mode
• Classifying enclaves created by the scalable Web server

Using Velocity Goals

• Measuring current velocities while migrating
• Work with low activity and high velocity goal not achieving its goal
• Selecting single, sysplex-wide velocity goal in a mixed-hardware sysplex

Operations - commands, policy switch, etc.

• "RESET jobname,QUIESCE" still allows non-swappable address space to be dispatched
• Disruption after switching to new policy
• Resetting discretionary work to better service class to cancel

I/O Priority Management

• Counting I/O disconnect time as I/O using time

Using Batch Initiator Management

• WLM initiator balancing in z/OS 1.4
• Recycling initiators getting ABEND822s
• Determining how many initiators WLM has started

Troubleshooting

• What happened to DB2 procedure? 

• System address spaces running in discretionary service class
• Overhead of goal mode over compatibility mode
• More information on WLM