Nagios Support Forum

Dear Friends

I have a host that has two different circuits. when a're running turns green and when he is in trouble turns red.

is it possible to place orange if I have a problem with running and another.

Thanks
Almir

So something like this?

Edit: Just noticed @tmcdonald's reply came in slightly before mine. I still think my advice below holds.

Dependencies are your answer. http://nagios.sourceforge.net/docs/3_0/ ... ncies.html

Dependencies might not be the answer here if OP is talking about having two possible routes for a ping to make. BPI would be a possibility if each link can be tested individually. But I would rather use a single check in this case.

I guess I need more information (or better translation) as to what the requester is asking for. My interpretation is that if the first link is down, he wants to go yellow, but only go red if the second link is also down.

yes....
but im want to see yellow when have one lin down + another one link up
is possible?

tmcdonald wrote:So something like this?

connection.png

My diagram isn't as nice as Trevor's, but here it is.

Assuming that it is GREEN if both links are ok, and RED if both links are down, what color do you want for the other two states?

I believe that's when he would warn a yellow WARNING state. So unless the individual lines are monitored and set as depends, BPI or a custom plugin is the answer here.

nono
Green - All link is UP
Yellow - one of the link is down
Orange - one of the link have ore than 80% of traffic usage (if is possible)
Red - all link is down

eloyd wrote:My diagram isn't as nice as Trevor's, but here it is.

Assuming that it is GREEN if both links are ok, and RED if both links are down, what color do you want for the other two states?

The colors are not arbitrary, but rather indications of the state of a host or service. Green is OK, Yellow is Warning, Red is Critical, and Orange is Unknown.

The only way you could get around this is if you had a custom plugin that output the proper status, but then you might run into problems where the Unknown status needs to be treated as a second Warning state.