GPA Manual

Running GPA

Guardian Performance Analyzer (GPA) Manual—135081

2-4

Running GPA

GPA does not require installation; it is configured to run from the installation subvolume

$SYSTEM.ZGPA. To run GPA, you simply set your current volume to the location of

your Measure structured data files and invoke GPA’s TACL macro

$SYSTEM.ZGPA.OBEYTUNE. When GPA completes the analysis, it generates the

output reports. You can direct the report to any printer device by suitably modifying the

OBEYTUNE file.

Tuning the System with PUPIN or SCFIN

In this final step of the procedure, you implement some of the GPA tuning

recommendations by running a PUP INPUT file, PUPIN, or an SCF INPUT file, SCFIN,

which GPA generates as part of its analysis. PUPIN/SCFIN contains commands for

redistributing primary disk process paths and resetting disk volume cache allocations. (If

you want to restore the system to its previous condition after you run PUPIN/SCFIN,

you can do so with another PUP or SCF INPUT file, PUPBAK/SCFBAK.) Other GPA

tuning recommendations are discussed in Section 4, Using GPA Information

, under

Implementing GPA Tuning Recommendations.

The GPA Procedure

To run GPA, follow the steps outlined below.

1. Establish the measurement window and data collection interval for running Measure

on the system, as described earlier.

2. Use the STATUS command to see if the Measure subsystem is currently running. If

it is, go directly to Step 3. Otherwise, start the Measure subsystem by entering:

1> MEASCOM

Figure 2-2. Code Characters in a GPA Text Statement Paragraph

Caution. Note that any changes in the disk primary CPU invoked by running GPA’s PUPIN or

SCFIN file are not permanent and will not be in effect after a cold load.

CDT .CDD

Keyword Code Characters Number of lines in the paragraph

Code

Characters

:TEXT SYSTEM CPUBALANCE F 07

CPU load balancing is FAIR:

OB of CC processors on this node appear to be out of

balance, considering the utilization level of this node

during the measurement period. As the overall utilization of

the node increases the requirement for "tighter" cpu

balancing will increase.

:END: