Collecting Performance Stats

Today I have been collecting some performance stats to investigate potential performance bottle necks.  Attached is the guide from HP which describes how to collect a group of performance stats using the service processor

Collecting_InServ_performance_data_using_Performance_Analyz

The process essentially runs the following command lines over a period of time:

statcpu -iter 1000 -d 15

statcmp -iter 1000 -d 15

statport -disk -rw -ni -iter 1000 -d 15

statport -host -rw -ni -both -iter 1000 -d 15

statvv -rw -ni -iter 1000 -d 15 Requested command: statvlun -rw -ni -iter 1000 -d statpd -rw -ni -devinfo -iter 1000 -d 15

statiscsisession -iter 1000 -d 15

statiscsi -fullcounts -counts -iter 1000 -d 15 Requested command: statfcoe -iter 1000 -d 15

 

The key option available to you is the number of iterations i.e. the number of times the command will run and the delay before executing those commands.  The above shows it will run 1000 times with a delay of 15 seconds between executions.

 

When it’s all finished you can send the results to HP, but you can also view the results yourself.  Hence this is quite a cool way to execute a bunch of performance commands quickly and easily without having to enter them all manually.  To view the results open a session to your service processor in a web browser, choose files down the left hand side, then choose files in the file explorer window,  serial number of the 3Par, Perfanalysis then choose the Zip with the correct date you want to look at.  Choose download and then you can view each of the stats collected in Internet Explorer.

 

Published by

2 thoughts on “Collecting Performance Stats

  1. Hey 3pardude!

    do you know where I can find a document discussing the 3par inservstats schema? want to generate reports with Crystal Reports and can’t find date/time to correlate with the TSECS in recent sample db.

    Thanks for any info

    Suzanne

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.