Graph for overall Hostgroup availability

[RebeccaIlene]

Hi Team,

We are looking to have a combined graph for all the hosts availability (ping) for a host group. Can you please suggest how this can be done?

I am aware that we can the Graph Explorer for combining the output for multiple graphs. However, we have around 2700 hosts in this host group and adding each host individually is not a feasible option.

Also, can you please help us to modify the check_ping output so that it only gives the RTA value?

Currently the check gives an output as:

PING CRITICAL - Packet loss = 20%, RTA = 1459.91 ms|rta=1459.906982ms;100.000000;150.000000;0.000000 pl=20%;80;100;0

We want to modify it so that we get only the value of RTA. For example output for above should be: 1459.91 ms.

Can you please confirm if this is possible?

Thanks,
Rebecca Murray

[jbrunkow]

It does not appear that the Graph Explorer in Nagios XI currently allows a Hostgroup to be selected as a Graph Object. However, that may be a worthwhile feature to request adding to future releases, considering use cases at scales such as yours.

You could use the cut command in Linux to filter out just the RTA value.

Code: Select all

check_ping [options] | cut -d ' ' -f 11 | cut -d ';' -f 1 | cut -d '=' -f 2

There may be a more precise way to pluck just that value from a body of text, but that is the first solution that comes to mind. Are you looking to modify the output so that it is only that value in text, or are you trying to make a graph using only that value?

Just in case you haven't come across this document yet, I found the following link to be useful in detailing the features of Nagios XI graphs.
https://assets.nagios.com/downloads/nag ... ios-XI.pdf

[RebeccaIlene]

Thanks for sharing that. However, yesterday we tried adding the hosts performance in the Graph Explorer and it won't take more than 95 inputs.
Can you please confirm if there is a limit to the number of inputs on graphs?

Is there any other way we can have all 2700 host performance on the same graph?

Yes, we are looking to have only the value if RTA generated as output and in the graph.

We want to monitor the ping performance for all the hosts to find out which are the top 10 with lest RTA and the lowest 10 with maximum RTA.

Will test the cut command provided and confirm if this works.

Regards,
Rebecca Murray

[jbrunkow]

Are you getting any error messages when you try to add more than 95? I do not believe that Nagios XI imposes a limit on how many inputs can be added, other than how much PHP is allocated to that machine. If you tail your Apache logs to see the most recent entries, you may be able to see more information about what the limitation might be.

If you are open to considering another solution, you may want to look at the linked article. You could possibly query the RRDs directly and accomplish something similar.
https://serverfault.com/questions/57468 ... mpty-graph