Comparison charts

Consider using a comparison chart to measure key aspects of your system performance before and after tuning changes have been made. A suggested chart is as follows:

Table 1. Comparison chart
Observations to make Run A Run B Run C Run D
DL/I transactions: Number        
DL/I transactions: Response        
VSAM transactions: Number        
VSAM transactions: Response        
Response times: DL/I        
Response times: VSAM        
Most heavily used transaction: Number        
Most heavily used transaction: Response        
Average-use transaction: Number        
Average-use transaction: Response        
Paging rate: System        
Paging rate: CICS®        
DSA virtual storage: Maximum        
DSA virtual storage: Average        
Tasks: Peak        
Tasks: At MXT        
Most heavily used DASD: Response        
Most heavily used DASD: Utilization        
Average-use DASD: Response        
Average-use DASD: Utilization        
CPU utilization      
This type of comparison chart requires the use of TPNS, RMF™, and CICS interval statistics running together for about 20 minutes, at a peak time for your system. It also requires you to identify the following items:
  • A representative selection of terminal-oriented DL/I transactions accessing DL/I databases
  • A representative selection of terminal-oriented transactions processing VSAM files
  • The most heavily used transaction
  • Two average-use nonterminal-oriented transactions writing data to intrapartition transient data destinations
  • The most heavily used volume in your system
  • A representative average-use volume in your system
To complete the comparison chart for each CICS run before and after a tuning change, you can obtain the figures from the following sources:
  • DL/I transactions: Identify a selection of terminal-oriented DL/I transactions accessing DL/I databases.
  • VSAM transactions: Identify a selection of terminal-oriented transactions processing VSAM files.
  • Response times: External response times are available from the TPNS terminal response time analysis report; internal response times are available from RMF. The “DL/I” subheading is the average response time calculated at the 99th percentile for the terminal-oriented DL/I transactions you have previously selected. The “VSAM” subheading is the average response time calculated at the 99th percentile for the terminal-oriented VSAM transactions you have previously selected.
  • Paging rate (system): The RMF paging activity report shows a figure for total system non-VIO non-swap page-ins added to the figure shown for the total system non-VIO non-swap page-outs. This figure is the total paging rate per second for the entire system.
  • Tasks: Transaction manager statistics (part of the CICS interval, end-of-day, and requested statistics). The “Peak” subheading is the figure shown for “Peak Number of Tasks” in the statistics. The “At MXT” subheading is the figure shown for “Number of Times at Max. Task” in the statistics.
  • Most heavily used DASD: The RMF direct access device activity report, which relates to the most heavily used volume in your system. The “Response” subheading is the figure shown in the “Avg. Resp. Time” column for the volume you have selected. The “Utilization” subheading is the figure shown in the “% Dev. Util.” column for that volume.
  • Average-use DASD: The RMF direct access device activity report, which relates to a representative average-use volume in your system. The “Response” subheading is the figure shown in the “Avg. Resp. Time” column for the volume you have selected. The “Utilization” subheading is the figure shown in the “% Dev. Util.” column for that volume.
  • Processor utilization: The RMF processor activity report.

This chart is most useful when comparing before-and-after changes in performance while you are tuning your CICS system.



dfht3c0092.html | Timestamp icon Last updated: Monday, 5 June 2017