beautypg.com

Capcustombenchmark(1m), Name, Synopsis – HP Matrix Operating Environment Software User Manual

Page 134: Description, Logging

background image

capcustombenchmark(1M)

NAME

capcustombenchmark -- Configure Capacity Advisor benchmark CPU scaling to normalize collected
data.

SYNOPSIS

Path on CMS:

/opt/vse/bin

capcustombenchmark

[{{-c-customonly} {{{{-f-inputfile}benchmark-file}}}}]

capcustombenchmark

[{{-b-bothCustomAndBuiltIn}

{{{{-f-inputfile}benchmark-file}}}}

]

capcustombenchmark

[-B-builtInOnly]

DESCRIPTION

The capcustombenchmark command allows you to configure the Capacity Advisor benchmark
database for scaling CPU utilization data when modeling the move of workloads from one system
to another. With capcustombenchmark, you can set the benchmark configuration to use only
custom benchmark data that you supply, use custom benchmark data first and fall back to benchmark
data provided by Capacity Advisor, or use only Capacity Advisor benchmark data. By default,
Capacity Advisor is configured to use only its own benchmark data.
This command is for use on servers running a CMS on HP-UX or Microsoft Windows.

Options

-c | -customOnly

Specifies that Capacity Advisor should use only benchmark
data that you supply. You must include {-f |
-inputFile

} with this option.

-b | -bothCustomAndBuiltIn

Specifies that Capacity Advisor should preferentially use
benchmark data that you supply, then use Capacity Advisor
benchmark data when custom data does not contain data
for the system(s) in question. You must include {-f |
-inputFile

} with this option.

-B | -builtInOnly

Specifies that Capacity Advisor should use only its own
benchmark data.

-f | -inputFile

Specifies that Capacity Advisor should use
benchmark-file

as input for normalization calculations.

LOGGING

Performance index results are calculated in the web application. To enable logging of performance
index results, add the following line to the /etc/opt/vse/vseprefs.props file:

benchmark.db.logresults=true

The results are logged to file: /var/opt/vse/logs/perfIndex.resultLog
In this file, you can see one line for each system for which a performance index has been calculated.
The format of each line is:

hostname=attribute:value,attribute:value,

....

For example,

hostname1=result:1.0 +- 1.0 : CLOCK_SPEED (1000.0),processor:,model:ia64 hp superdome server SD32A,Core
count:4,Database:BASELINE
hostname2=result:1.0 +- 1.0 : CLOCK_SPEED (1000.0),processor:Intel(R) Itanium(R) 2 ,model:,Core
count:16,Database:BASELINE

The attributes displayed are:

134

Command reference