Welcome to the beta test program for NmonCheck.
This beta version is meant for people who have some familiarity with our
SarCheck analysis tool. The primary difference between NmonCheck and
SarCheck is that SarCheck starts with sar data and then collects the other
data that it needs. NmonCheck starts with the CSV output from nmon.
Because nmon and sar do not monitor the same things or monitor them with
the same precision, there may be some differences in the output of the two
tools. If the values of various metrics are very close to thresholds used
by the rule sets of the two programs, there may be differences in the
recommendations. We don't expect any major differences but some are
inevitable.
The basic differences between SarCheck and NmonCheck:
Location of the software /opt/sarcheck /opt/nmoncheck
Name of the analysis pgm analyze nanalyze
The front end script complete still needs work
Add'l data collected by our agent & ps our agent
Name of the agent scaixagent ncagent
The script for the agent complete seems to work correctly
Location of agent data /opt/sarcheck/ps /opt/nmoncheck/nc
Switches available see the documentation see the documentation
Status of HTML output complete seems to work correctly
Some information about nmon:
The nmon utility was written by IBM's Nigel Griffiths in the mid-90's and
has been available to the public without IBM's official support for about
10 years. The brilliance of nmon is its simple output and low overhead.
IBM recently embraced nmon and incorporated it into the topas utility. It
is supposed to work the way it always has. The nmon example here uses the
original nmon program. The -f switch saves nmon's data to a file. The -s
switch specifies the number of seconds between samples, and the -c switch
is a count of the number of samples which will be collected. The NmonCheck
software is designed to look at data which is collected in intervals of 5
to 60 minutes. Here is an example of a crontab entry used to collect data
between 9:00AM and 5:00PM:
0 9 * * * /usr/bin/nmon -f -s 600 -c 49
Additional data can be collected by our agent. To collect data for the
same time period, use the following crontab entries:
0,10,20,30,40,50 9-16 * * * /opt/nmoncheck/bin/nc1
0 17 * * * /opt/nmoncheck/bin/nc1
For 24x7 monitoring, use these crontab entries:
0 0 * * * /usr/bin/nmon -f -s 600 -c 144
0,10,20,30,40,50 * * * * /opt/nmoncheck/bin/nc1
Here is an example of a command used to run NmonCheck. In this example,
we're going to create the sample report seen on the aptitune.com website.
In this example, the /opt/nmoncheck/bin directory is in the PATH:
nanalyze -png -html -t -dtoo sample.nmon > samplereport.html
To see all the switches and understand how to use them, just run the
nanalyze program with the -h switch, and you'll probably want to pipe the
output to more. We've tried to make the switches the same as the SarCheck
switches and there are only a few exceptions.
This is very much a "Version 1" product right now, and we're adding new
features as fast as we can. Some of these limitations just indicate how far
along we are, others are due to a lack of data to analyze or a lack of clear
documentation. Here is the list of some of the things that NmonCheck
doesn't do yet:
1. NmonCheck doesn't analyze multiple days of data in one report yet
2. Some of the text might still mention sar or ps
3. NmonCheck can currently analyze only the first 150 disks
4. Disk service time stats are not available for older AIX systems
5. We're still looking for the disk queueing stats
6. NmonCheck uses nmon's PROC and TOP data instead of ps -elf statistics
7. The front end script (in SarCheck it's called 'sarcheck') isn't
finished yet. You'll need to use the command line.
Copyright 2009 Aptitune Corporation. All rights reserved.