NCPA Unknown status after changing binding address

[CoMAdmin]

I have NCPA deployed on multiple nodes and I noticed that an increasing number of them are showing a status of "UNKNOWN: An error occured connecting to API. (Connection error: '[Errno 111] Connection refused')"

This is for all services related to specific nodes. Its happening to all OS's: Server 2008R2, 2012, 2016. I was testing to see if this could have to do with the IP address that I am entering for the listener. It turns out that if I leave the IP address as the default 0.0.0.0, the status goes back to normal and I am able to get full stats on these services. The moment that I enter the IP address of the Nagios server in the "allowed_sources" field and restart the NCPA Listener service, I start to get the same errors.

I do have many servers working correctly with the IP address of the Nagios server in that field. I also tried to copy that cfg file from a working node to one of the trouble nodes but no difference in the behavior.


The goal is for all servers that communicate with Nagios XI using NCPA to only listen to the Nagios box on that NCPA port.


Any help on this would be appreciated.

[benjaminsmith]

HI,

I'm not 100% certain why this is working for some systems and not for others, but you'll want to set this up using the allowed_hosts option to limit access.

allowed_hosts A comma-separated list of ip addresses of hosts that would be allowed to connect to NCPA.

Let me know if that's a suitable workaround, unless you are using the allowed_sources option as well to restrict GUI access as well.