Nagios 3: ping service

[fff7878]

Hello all,
I've just installed Nagios 3, ndo2db and centreon 2. I am testing it to monitor about 100 servers (and 15 switches).
I've search around on this forum and the internet but can't find a clear answer about the "check_host_alive" command and defining a PING service associated to it.
So, what is clearly recommanded:
- Check_host_alive pinging only once the server and the ping service to have some graphs
- Check_host_alive with 3 or 4 pings being sent and no service
- Another solution...

I also have some misc questions:
- For the switches and other network equipments, do you define a PING service or just SNMP queries?
- in order to monitor disk usage, load average and CPU usage, is it better to use the check provided by centreon using SNMP or to use NRPE Nagios plugin?

Thansk in advance for your answer and your help,

FFF

[fff7878]

A little UP...

[naparuba]

Hi

Hello all,
I've just installed Nagios 3, ndo2db and centreon 2. I am testing it to monitor about 100 servers (and 15 switches).

Good choice ; )

I've search around on this forum and the internet but can't find a clear answer about the "check_host_alive" command and defining a PING service associated to it.
So, what is clearly recommanded:
- Check_host_alive pinging only once the server and the ping service to have some graphs
- Check_host_alive with 3 or 4 pings being sent and no service
- Another solution...
It's recommended to make one ping by check_host_alive. But you have to put a "Max Check Attempts" so that with just one test you do not receive notification. One ping lost is not important, one another lost one minute after is important: you lose the connection, not just a packet. At this moment, you want to be notified. You've go one minute delay, but you will received only REAL alerts.


I also have some misc questions:
- For the switches and other network equipments, do you define a PING service or just SNMP queries?
The Ping must be for the host check, not a service. So
ping (check_host_alive) = host check
snmp (interface up down, load, etc)= service check


- in order to monitor disk usage, load average and CPU usage, is it better to use the check provided by centreon using SNMP or to use NRPE Nagios plugin?
If the information into SNMP is enough for you, no not install NRPE.

[fff7878]

Thanks a lot for the reply!
So:
- I'll remove the PIN check as service and modify the hosts checks as you recommend
- Remove the ping to the switches and use SNMP
- Use SNMP for the checks... as long as it's enough!

One last question:
- for the switches, Do I also use check_host alive command on the IP address of the switch or another check?

Thanks,
FFF

[jeeva]

I found this site for ping test..I test my ping test here http://www.whoisxy.com/ping.aspx at free of cost...