NagiosXI Server Migration Problem

[nscott]

Hi Aurelien,

That unit problem is has been narrowed down in the past to be a pnp template issue. Can you post the contents of your /usr/local/nagios/share/pnp/templates/check_xi_mrtgtraf.php?

[srrhd]

Here it is:

Code: Select all

[root@NagiosXI scripts]# cat /usr/local/nagios/share/pnp/templates/check_xi_service_mrtgtraf.php
<?php
#
# Copyright (c) 2006-2008 Joerg Linge (http://www.pnp4nagios.org)
# Plugin: check_snmp_int.pl (COUNTER)
# $Id: check_xi_service_mrtgtraf.php 760 2011-08-03 17:09:09Z mguthrie $
#
#
$opt[1] = " --vertical-label \"Traffic $UNIT[1]\" -E --title \"Interface Traffic for $hostname / $servicedesc\" ";
$def[1] = "DEF:var1=$rrdfile:$DS[1]:AVERAGE " ;
$def[1] .= "DEF:var2=$rrdfile:$DS[2]:AVERAGE " ;
$def[1] .= "CDEF:real1=var1,1,* " ;
$def[1] .= "CDEF:real2=var2,1,* " ;
$def[1] .= "LINE1:real1#003300:\"in  \" " ;
$def[1] .= "GPRINT:real1:LAST:\"%7.2lf $UNIT[1] last\" " ;
$def[1] .= "GPRINT:real1:AVERAGE:\"%7.2lf $UNIT[1] avg\" " ;
$def[1] .= "GPRINT:real1:MAX:\"%7.2lf $UNIT[1] max\\n\" " ;
$def[1] .= "LINE1:real2#00ff00:\"out \" " ;
$def[1] .= "GPRINT:real2:LAST:\"%7.2lf $UNIT[1] last\" " ;
$def[1] .= "GPRINT:real2:AVERAGE:\"%7.2lf $UNIT[1] avg\" " ;
$def[1] .= "GPRINT:real2:MAX:\"%7.2lf $UNIT[1] max\\n\" ";
if($NAGIOS_TIMET != ""){
    $def[1] .= "VRULE:".$NAGIOS_TIMET."#000000:\"Last Service Check \\n\" ";
}
if($NAGIOS_LASTHOSTDOWN != ""){
    $def[1] .= "VRULE:".$NAGIOS_LASTHOSTDOWN."#FF0000:\"Last Host Down\\n\" ";
}
?>

For information we have never modified those files.

Aurelien.

[srrhd]

Any ideas ?

[mguthrie]

Can you post a sample of the plugin output with performance data, as well as an XML file for one of these services in /usr/local/nagios/share/perfdata/<host_name>/.

[srrhd]

Comand line output:

Code: Select all

[root@NagiosXI]# sudo -u nagios /usr/local/nagios/libexec/check_rrdtraf -f /var/lib/mrtg/IP_ADDRESS_INT.rrd -w 50,50 -c 70,70 -l M
OK - Current BW in: 0Mbps Out: 0Mbps|in=.000008Mb/s;50;70 out=.000008Mb/s;50;70

Below the screen shots from XI interface. For the XML file i send it to you via Private Message.

[mguthrie]

Grr, I was hoping this answer would show up in the template, but I think from here the issue may be in rrdtool somehow. As you mentioned earlier, this issue showed up in another thread, but from what I can tell it may not have been resolved. I'm wondering it this is an issue of too many decimal places, or rrdtool attempting to convert the number into a different unit of measurement. I'm a little bit stumped at the moment, I'll ask around and see if anyone else has any ideas.

<UNIT>Mb/s</UNIT>
<ACT>.000008</ACT>

[srrhd]

OK... I try to make my boss wait a bit more...

I go on my investigation too and hope for your return.

[nscott]

Aurelien,

Those numbers look strange, the .000008Mb/s would translate to ~8.3 b/s, and since pnp, but that .000008 is definitely getting rounded by the check_rrdtraf check, what is returned when you do use the '-l M' on that check?

[srrhd]

Hi Nicholas,

I am not sure what you are asking me for as we already use the "-l M" option.

So, I tryied without :

Code: Select all

[root@NagiosXI nagiosxi]# sudo -u nagios /usr/local/nagios/libexec/check_rrdtraf -f /var/lib/mrtg/IP_ADDRESS_INT.rrd  -w 50,50 -c 70,70
OK - Current BW in: 8.00bps Out: 24.00bps|in=8.000000b/s;50;70 out=24.000000b/s;50;70

Looking the interface statistics directly on the equipement gives:

Code: Select all

  5 minute input rate 9407000 bits/sec, 1297 packets/sec
  5 minute output rate 1747000 bits/sec, 870 packets/sec

Which seems to implies that there is an error by 1000000...

For information this morning i upgrade the server to 2011R2.2 and the problem is going on.

Aurelien.

[nscott]

Aurelien,

Sorry about that, I re-read that sentence I wrote before and you did the correct thing, I had dropped some words from that sentence .

Can we check to see if the MRTG RRD is seeing the same numbers as well? Can you located the RRD file under /var/lib/mrtg/ and run an:

rrdtool fetch <filename> AVERAGE -s '-1d'

Do those numbers look right? If they don't then perhaps MRTG is polling something wrong or is doing some math on the numbers.