Hi, I have discussed this issue on IRC but haven't got a resolution.
Issue summary: My setup is such that host checks are run from the same host, and the host then sends check results to master. This means that when the host machine itself is down, the master keeps showing the host as UP, which was the last status before the host went down. Setup summary: I am using central config on master, with the intent that checks for a Host (client) will be locally executed from the Host itself. For this the suggestion on IRC was to define the Hosts inside Zone named as the Host itself. And then also have endpoint defined inside the zone as the host itself. I searched through icinga mailing list archives and got to know that checks for a respective Host object is run from Endpoint(s) defined in the Host's Zone. So this is the setup that is working for me. Setup detail: Let my master host be master101, and the Host/client I want to monitor be host101. Then on master, I have an endpoint and a zone named "host101" and an endpoint and zone named "master101". Also i have defined a global zone. I then have a file in this zone placed at /etc/icinga2/zones.d/host101/hosts.conf. This file only contains a Host object called "host101". Then there is a folder /etc/icinga2/zones.d/global/applyservices.conf which has generic service checks define. On host101, i have the same zones.conf as master, and so there is a zone called "host101" with endpoint "host101", another zone called "master101" with endpoint "master101" and a global zone. Now, the host101 can sync the zones.d/host101/ folder from master and is able to run all checks correctly. Setup config files: available on request. But i think the setup is quite simple. Problem description: Now I have understood from mailinglist archives that check for a Host is run from Endpoint defined in the Zone of the Host object. In my case, I have deliberately made sure that all of Host, Zone and Endpoint fall on the same physical server, and so all checks get locally executed. The icingaweb2 on master101, correctly shows realtime status of these service checks for host101. But this has a problem. The host check is "hostalive" and is also running locally. So if the server goes down, hostalive doesn't get executed and master doesn't know if the host is down, and keeps showing as UP. So, I want to know if I can run only the host check from, lets say, the master and correctly associate it with the Host? So that when host101 is down, it gets correctly marked on icingaweb2 and also, all checks start to show UNKOWN? Or is there any alternative config that i should try? Thanks a lot. I have seen a similar problem https://lists.icinga.org/pipermail/icinga-users/2015-May/009519.html with no solution. -indrajeet
_______________________________________________ icinga-users mailing list icinga-users@lists.icinga.org https://lists.icinga.org/mailman/listinfo/icinga-users
