I have two models, both using Openstack instances. Both are in the same Openstack network.
The model with Nagios is connected to nrpe via cross model relation.
When I look at nrpe.cfg, the public (i.e. floating) IP address of the Nagios host is specified as allowed. When I check the host definitions in Nagios, the private (non-floating) address is used. This means that Nagios doesn't have permission to connect to NRPE because it uses it's private address.
This is intermittent - somehow one iteration of config-changed supplied private, and then after reboots etc the other address was used and we were unable to reach nrpe.
This has also been observed with manual machines outside Openstack instances, where the nagios host has multiple addresses and is attempting to use one not listed to connect to the manually added machine.
I have two models, both using Openstack instances. Both are in the same Openstack network.
The model with Nagios is connected to nrpe via cross model relation.
When I look at nrpe.cfg, the public (i.e. floating) IP address of the Nagios host is specified as allowed. When I check the host definitions in Nagios, the private (non-floating) address is used. This means that Nagios doesn't have permission to connect to NRPE because it uses it's private address.
This is intermittent - somehow one iteration of config-changed supplied private, and then after reboots etc the other address was used and we were unable to reach nrpe.
This has also been observed with manual machines outside Openstack instances, where the nagios host has multiple addresses and is attempting to use one not listed to connect to the manually added machine.
Imported from Launchpad using lp2gh.
date created: 2019-06-13T06:02:03Z
owner: xavpaice
assignee: None
duplicate of: #50
the launchpad url