Nagios XI SNMP collection on discovery mode

[slansing]

The reason for installing certain agents is two fold, for active plugin checks you need to have a way to remotely execute plugins, then return the output to Nagios is a way readable to Nagios, this is where NRPE or NCPA come in. For passive results, such as those retrieved with NRDS or NSCA there needs to be a way to execute the plugins at relatively the same time, and then push the results to Nagios. You can use check_by_ssh to execute plugins and return the output as well as sreinhardt mentioned and this is covered here:

http://assets.nagios.com/downloads/nagi ... ng_SSH.pdf

Certain checks allow you to specify a host address, and query specific information with the local nagios plugin on the nagios server, checks such as check_ping, check_http, mysql, etc.

[rashidaq]

thank you guys for the quick response

I have been told by management for not to install agents

Maybe change/configure the snmp agents for those to be monitored process/application, maybe...

I learned alot

thank you

Ras

[slansing]

Okay, I'd recommend you look into the different protocols a bit more before actually setting up service checks as you set a ton of NRPE checks up, without following our documentation to install the agent "which is in the first step of the NRPE/Linux Server/Windows Server wizards.. Let us know what you decide to do. It looks like you had started to configure SNMP checks against your Windows/*nix hosts, but keep in mind for active SNMP checks you do need port 161 open.

[rashidaq]

The SNmpwalk and snmpget I am doing on linux with some application running
The applications need to be monitored as mem usage, cpu usage... from Nagios
but cannot be done.
I can get general SNMP results such as linux system perfomance thru Nagios but not specific
application performance

thanks for your help

Ras

[abrist]

Have you tried using check_snmp_process.pl ?

[rashidaq]

no, I havent yet

as reminder these applications(to be monitored) running on the hosts do not have MIBS/OID s

so how does that work? check_snmp_process.pl ?


thanks

[abrist]

If the host is running snmpd, you should have some oids/mibs to check, it anything the standard SNMP MIB 2 should be available with the daemon.

[lmiltchev]

You may use something similar to this:

Code: Select all

/usr/local/nagios/libexec/check_snmp_process.pl -H <remote box ip>  -C <community> --v2c -n 'explorer.exe'
1 process matching explorer.exe (> 0)

[rashidaq]

thanks
but two more questions

there are two processes that uses java, those are red5 and tomcat
but when I do
[root@nagios ~]# /usr/local/nagios/libexec/check_snmp_process.pl -H 192.168.156.183 -C DCSZONEA --v2c -n 'java'
2 process matching java (> 0)
[root@nagios ~]# /usr/local/nagios/libexec/check_snmp_process.pl -H 192.168.156.183 -C DCSZONEA --v2c -n 'red5'
No process matching red5 found : CRITICAL

so how do I get this done for red5 or tomcat? and not its java processes..? please

also is there a plugin for nagios XI for graphical output?

[scottwilkerson]

You likely will need to use the -A parameter, like

Code: Select all

/usr/local/nagios/libexec/check_snmp_process.pl -H 192.168.156.183 -C DCSZONEA --v2c -n 'java' -A 'red5'