check_snmp, performance graph and graph explorer discrepancy

[tmcdonald]

I spoke to the developers and it looks like the graph is showing a change over time instead of the value itself for whatever reason. I don't have a fix yet but I wanted to keep you updated.

[iptribe]

I spoke to the developers and it looks like the graph is showing a change over time instead of the value itself for whatever reason. I don't have a fix yet but I wanted to keep you updated.

Does it have something to do with the letter "c" in the output shown in the group instead of numbers only? Because even if it is showing a change over time, it is not supposed to be showing decimal numbers as the values are from 0 and up.

[tgriep]

The letter "c" on the output lets the system know that the data is a counter.
There is a formula that the data gets processed through and what you are seeing is rounding errors from that formula.
We are still looking at the issue and we think that a graph template for SNMP counters will have to be created.

If you do remove the letter "c" from the output, it will graph correctly.

[iptribe]

The letter "c" on the output lets the system know that the data is a counter.
There is a formula that the data gets processed through and what you are seeing is rounding errors from that formula.
We are still looking at the issue and we think that a graph template for SNMP counters will have to be created.

If you do remove the letter "c" from the output, it will graph correctly.

I shall wait for the new template for SNMP counters then. Tried to edit the existing plugin check_snmp, however, it is in binary format...

[jdalrymple]

Binary, sure, but the source is readily available:
https://github.com/nagios-plugins/nagio ... eck_snmp.c

I think lines 400-411 are what's gettin ya

[iptribe]

Binary, sure, but the source is readily available:
https://github.com/nagios-plugins/nagio ... eck_snmp.c

I think lines 400-411 are what's gettin ya

Thanks jdalrymple. I believe this is the same one that I have. Which does not show the right value in the graph...

[ssax]

It is the same one, jdalrymple was just offering you the code in case you wanted to try to fix it and recompile it.

I've edited the code to remove the "c" from the sections that jdalrymple mentioned (see attached).

*** Note: this does change it for all counters being returned

check_snmp.zip

Here is how you compile it:

Code: Select all

cd /tmp
wget http://nagios-plugins.org/download/nagios-plugins-2.0.3.tar.gz
tar zxf nagios-plugins-2.0.3.tar.gz

Unzip the attachment and replace /tmp/nagios-plugins-2.0.3/plugins/check_snmp.c then:

Code: Select all

cd /tmp/nagios-plugins-2.0.3
./configure --with-nagios-user=nagios --with-nagios-group=nagios 
make
mv /usr/local/nagios/libexec/check_snmp /usr/local/nagios/libexec/check_snmp.old
cp /tmp/nagios-plugins-2.0.3/plugins/check_snmp /usr/local/nagios/libexec/

[iptribe]

It is the same one, jdalrymple was just offering you the code in case you wanted to try to fix it and recompile it.

I've edited the code to remove the "c" from the sections that jdalrymple mentioned (see attached).

*** Note: this does change it for all counters being returned

check_snmp.zip

Here is how you compile it:

Code: Select all

cd /tmp
wget http://nagios-plugins.org/download/nagios-plugins-2.0.3.tar.gz
tar zxf nagios-plugins-2.0.3.tar.gz

Unzip the attachment and replace /tmp/nagios-plugins-2.0.3/plugins/check_snmp.c then:

Code: Select all

cd /tmp/nagios-plugins-2.0.3
./configure --with-nagios-user=nagios --with-nagios-group=nagios 
make
mv /usr/local/nagios/libexec/check_snmp /usr/local/nagios/libexec/check_snmp.old
cp /tmp/nagios-plugins-2.0.3/plugins/check_snmp /usr/local/nagios/libexec/

I did as you instructed, but it is still showing the same values in the graph with "c". Do I have to delete the existing rrd and xml files for that service so it to generates a new one?

[tgriep]

I have the graph template done that should fix the graphing issue with monitoring counters with SNMP. Here is how to install it.

In Core Config Manager
Create a copy of the check_snmp command and call it check_snmp_gauge

Edit the Command Name to check_snmp_gauge
define command {
command_name check_snmp_gauge
command_line $USER1$/check_snmp -H $HOSTADDRESS$ $ARG1$
}

Then edit the service check to use the check_snmp_gauge command like below

Code: Select all

define service {
	host_name			xxx.xxx.xxx.xxx
	service_description		Test SNMP
	check_command			check_snmp_gauge! -P 2c -C public -o  1.3.6.1.2.1.4.31.1.1.35.1 -w 100 -c 110 -l test!!!!!!!
	max_check_attempts		5
	check_interval			5
	retry_interval			1
	check_period			xi_timeperiod_24x7
	notification_interval		60
	notification_period		xi_timeperiod_24x7
	contacts			nagiosadmin
	register			1
	}

The graph template is called check_snmp_gauge.cfg and it needs to be uploaded to this folder in the XI system /usr/local/nagios/etc/pnp/check_commands/
Changed the permissions of the check_snmp_gauge.cfg file to the following
chown apache.nagios check_snmp_gauge.cfg
chmod 666 check_snmp_gauge.cfg

Delete the .rrd and .xml files for that service. They are in this folder under the hostname
/usr/local/nagios/share/perfdata/<HostName>/

After 15 minutes, the graph should be populated with the data.

[iptribe]

I have the graph template done that should fix the graphing issue with monitoring counters with SNMP. Here is how to install it.

In Core Config Manager
Create a copy of the check_snmp command and call it check_snmp_gauge

Edit the Command Name to check_snmp_gauge
define command {
command_name check_snmp_gauge
command_line $USER1$/check_snmp -H $HOSTADDRESS$ $ARG1$
}

Then edit the service check to use the check_snmp_gauge command like below

Code: Select all

define service {
	host_name			xxx.xxx.xxx.xxx
	service_description		Test SNMP
	check_command			check_snmp_gauge! -P 2c -C public -o  1.3.6.1.2.1.4.31.1.1.35.1 -w 100 -c 110 -l test!!!!!!!
	max_check_attempts		5
	check_interval			5
	retry_interval			1
	check_period			xi_timeperiod_24x7
	notification_interval		60
	notification_period		xi_timeperiod_24x7
	contacts			nagiosadmin
	register			1
	}

The graph template is called check_snmp_gauge.cfg and it needs to be uploaded to this folder in the XI system /usr/local/nagios/etc/pnp/check_commands/
Changed the permissions of the check_snmp_gauge.cfg file to the following
chown apache.nagios check_snmp_gauge.cfg
chmod 666 check_snmp_gauge.cfg

Delete the .rrd and .xml files for that service. They are in this folder under the hostname
/usr/local/nagios/share/perfdata/<HostName>/

After 15 minutes, the graph should be populated with the data.

Hi,

I did as u instructed. However, there is no change in the graph.

Command:

2.PNG

Service:

1.PNG

$ARG1$ : -P 2c -C public -o globalStatActiveCalls.summaryPeriod1Current5Minutes -m /usr/share/snmp/mibs/perimeta.mib -c 10000 -l ActiveCalls

Graph Template:
[xxx@NFVI-COMMON-NAGIOS-01 ~]$ ls -l /usr/local/nagios/etc/pnp/check_commands/ |grep snmp
-rw-rw-rw- 1 apache nagios 776 Jul 27 10:10 check_snmp_gauge.cfg


Graph:

3.PNG

Deleted the rrd and xml files too. Did I miss anything?