Cheshire Cat Computing

If you are monitoring a multi-CPU server, I suspect you would want to show the individual CPU percentages, plus an average.

You can do this using Userdefined graphs.

First, create four targets, each one graphing a single CPU. I don't know what agent you are using for this, probably SNMP? Anyway, all of these will probably be 'noo' option graphs.

Next, create a userdefined graph over tham all:

routers.cgi*Graph[cpu1]: allcpu average nototal
routers.cgi*Graph[cpu2]: allcpu
routers.cgi*Graph[cpu3]: allcpu
routers.cgi*Graph[cpu4]: allcpu
routers.cgi*Title[allcpu]: Combined CPU graph

Fnially, you may want to hide the individual CPU graphs:

routers.cgi*InMenu[cpu1]: no
...etc...

There are examples of this in the HOWTO and EXAMPLES files, and also in the online demo .cfg files.

Steve

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

i'm a bit confused with the average graphs you are talking about.
I have a 4 cpu system now that i monitor via snmp.
It graphs the targets separately.
I'd like to keep those and add the average graph.
I've scoured the examples and can't find an example that makes sense to me.

Could you point me in the right direction?

Read the HOWTO documentation about userdefined graphs. You need to make a userdefined that combines the 4 and adds an average line.

The EXAMPLES file contains examples of how to do this.

Also the online demo will show you the .cfg files for this sort of userdefined graph (see the mail server example).

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Hello Steve,

Sorry for my bad English and stupid questions.
MRTG is my memory of end of training course and I have to deploy on 1000 machines.
I still having pb to monitoring the systems with 2 CPU and 4 CPU.
You suggest me to create four targets, each one graphing a single CPU but I don’t know how I can do it? I copy my conf files from Internet and modified them to adapt to my configuration.
I tested to get OID with mib browser Getif but the result value is wrong.
I found this OID value with getif .1.3.6.1.2.1.25.3.3.1.2.1.1 for first CPU .1.3.6.1.2.1.25.3.3.1.2.1.2 for second but it’s not ok, so I put # before this target in my conf file.

### CPU Utilisation

Target[10.11.231.1_cpu]:.1.3.6.1.4.1.311.1.1.3.1.1.2.1.3.1.48&.1.3.6.1.4.1.311.1.1.3.1.1.2.1.3.1.49:public@10.11.231.1
#Target[10.11.231.1_cpu]:.1.3.6.1.2.1.25.3.3.1.2.1&.1.3.6.1.2.1.25.3.3.1.2.2:public@10.11.231.1
routers.cgi*Graph[10.11.231.1_cpu1]: allcpu average nototal
routers.cgi*Graph[10.11.231.1_cpu2]: allcpu
routers.cgi*Title[10.11.231.1_cpu]: Combined CPU graph
MaxBytes[10.11.231.1_cpu]: 100
Title[10.11.231.1_cpu]: CPU Utilization (average)
PageTop[10.11.231.1_cpu]: <H1>10.11.231.1 - Utilisation CPU (%)</H1>
<TABLE>
<TR><TD>System:</TD> <TD>10.11.231.1 </TD></TR>
<TR><TD>Maintainer:</TD> <TD>Administrateur </TD></TR>
<TR><TD>Max Value:</TD> <TD>100 %</TD></TR>
<TR><TD>Ip:</TD> <TD>10.11.231.1 (10.11.231.1.domain)</TD></TR>
</TABLE>
Options[10.11.231.1_cpu]: gauge,nopercent,
YLegend[10.11.231.1_cpu]: Utilisation CPU (%)
ShortLegend[10.11.231.1_cpu]: %
Colours[10.11.231.1_cpu]: GREEN#00eb0c,BLUE#1000ff,BLUE#1000ff,VIOLET#ff00ff
Legend1[10.11.231.1_cpu]:&nbsp;CPU Utilisation
Legend2[10.11.231.1_cpu]:&nbsp;CPU Utilisation
Legend3[10.11.231.1_cpu]:&nbsp;CPU Utilisation
Legend4[10.11.231.1_cpu]:&nbsp;CPU Utilisation
LegendI[10.11.231.1_cpu]:&nbsp;CPU 1
Legendo[10.11.231.1_cpu]:&nbsp;CPU 2

[/img]

I've also had some issues monitoring cpu's on windows servers.
on my quad cpu box i have had to use:
.1.3.6.1.2.1.25.3.3.1.2.1
.1.3.6.1.2.1.25.3.3.1.2.2
.1.3.6.1.2.1.25.3.3.1.2.3
.1.3.6.1.2.1.25.3.3.1.2.4
on my single cpu boxes some have worked with:

.1.3.6.1.2.1.25.3.3.1.2.1
while others have had to use:
.1.3.6.1.2.1.25.3.3.1.2.3
Weird stuff....but this is microsoft were talking about.
I have one dual cpu box that i cannot get any values from on cpu.
ive tried:.1.3.6.1.2.1.25.3.3.1.2.1 - 4 and .1.3.6.1.2.1.25.3.3.1.2.40 - 44

no luck.
I hope this helps.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11