Views-Service Details shows netsnmp error for all devices

[mfernacz]

I do not understand. The other device check should not be affected by the other host being unreachable. They are different IP addresses. Like I stated before it seems that Check Load and Check Memory are returning the same values for both devices. If I add another device it will show the same results.

[sreinhardt]

I think more than anything we are a little confused on exactly what you are attempting to do here and more-so prove by pulling the cable on the .11 device. Are you looking to use nrds to monitor remote systems that have snmp enabled but are not normally accessible via the nagios server directly? Could you give us an overview of what you are looking to implement and how you see it working? It does seem odd that the memory and load are coming back with the same results, could you also post the nrds config commands for those please.

[mfernacz]

I am trying to monitor Ethernet switches. I would like to gather those stats and generate reports for this network simulation network we are building. In Views->Service Details where it shows the status of the devices is nice, as you can see it is misleading. If I cause one device to be unreachable, I see no reason or why it would affect other devices in Views->Service Details. I would also like to have notification if something goes wrong and send out email. I have not played with that part yet.

command[Check Load]=/usr/local/nagios/libexec/check_load -w 15,10,5 -c 30,25,20
command[Check Memory]=/usr/local/nagios/libexec/check_snmp_mem_mrv.pl -H Core1_10_Odd -C public -w 60,60 -c 90,90 -f

command[Check Load]=/usr/local/nagios/libexec/check_load -w 15,10,5 -c 30,25,20
command[Check Memory]=/usr/local/nagios/libexec/check_snmp_mem_mrv.pl -H Core2_11_Even -C public -w 60,60 -c 90,90 -f


[root@localhost libexec]# crontab -u nagios -l
*/5 * * * * /usr/local/nrdp/clients/nrds/nrds.pl -H Core1_10_Odd 2>&1

*/5 * * * * /usr/local/nrdp/clients/nrds/nrds.pl -H Core2_11_Even 2>&1

[root@localhost libexec]# nmap -p 161 Core1_10_Odd

Starting Nmap 5.51 ( http://nmap.org ) at 2013-08-04 01:22 EDT
Nmap scan report for Core1_10_Odd (192.168.58.10)
Host is up (0.0013s latency).
PORT STATE SERVICE
161/tcp filtered snmp

Nmap done: 1 IP address (1 host up) scanned in 0.05 seconds
You have mail in /var/spool/mail/root
[root@localhost libexec]# nmap -p 161 Core2_11_Even

Starting Nmap 5.51 ( http://nmap.org ) at 2013-08-04 01:22 EDT
Nmap scan report for Core2_11_Even (192.168.58.11)
Host is up (0.0014s latency).
PORT STATE SERVICE
161/tcp filtered snmp

Nmap done: 1 IP address (1 host up) scanned in 0.05 seconds

[sreinhardt]

OK, I totally get wanting to monitor network switches, however it is still not clear to me why you are choosing to use nrds to do this. The only reason I can think to do it your present way is if this is a remote network from the nagios instance, which has not been explained. Not that it shouldn't work this way as well. Are you in a situation where each of the two devices you are running nrds on, are behind a separate switch, and you need to use those devices specifically to monitor those switches, such that XI could not directly run the snmp queries?

[mfernacz]

I was using NRDS because it was one of the videos I watched and assume that is what was needed to do for this to work. If there is a another way of doing this that works better I am open.

[abrist]

The most common way to monitor switches is through snmp. Have you tried the switch and router wizard?

[mfernacz]

Yes. That is how I added the 2 switches, but my boss also wants to monitor load and memory. That is why I used the NRDS.

[abrist]

What type of switches are these? You may be able to monitor those other metrics through snmp.

[sreinhardt]

Considering you can do the switch and router wizard via nagios off the bat, lets run those memory and processor checks straight from xi as well. I would suggest adding the check_mem and check_load to the correct libexec dir on the nagios server and define your commands similar to below:

Command Definitions:
command[Check SNMP Load]=/usr/local/nagios/libexec/check_snmp_load.pl -H $HOSTADDRESS$ -C $ARG1$ -w $ARG2$ -c $ARG3$ -T netsl -f
command[Check SNMP Memory]=/usr/local/nagios/libexec/check_snmp_mem_mrv.pl -H $HOSTADDRESS$ -C $ARG1$ -w $ARG2$ -c $ARG3$ -f

Service Definitions:
Setup a new service check for the hosts that you need this on, and set the argument values as follows
$ARG1$ = Community String
$ARG2$ = Warning Values
$ARG3$ = Critical Values

This should work out the issues you are having, if it does not or if you continue to have other issues related to this please let us know.

[mfernacz]

Thanks. These are MRV Ethernet switches.

So you are saying do not use NRDS and just use the scripts in /usr/local/nagios/libexec?

I am not sure how to add them to xi, so I will check it out. Is there any document or video that explains how to set it up?