Hello,
I use ansible to configure several servers and use “delegate_to” to write out their nagios configs on the nagios server. If the config is changed a notify handler is fired like below:
That may be a bug, as handlers are only supposed to trigger once of course despite the number of notifications sent to it. Could you open an issue on github for this?
Thanks!
Are you repeating this task among several plays? Handlers will fire once
per play, not per playbook.
It’s called once in one play in the “common” role i.e. once for each server that matches the “–limit” option when I run ansible-playbook.
e.g.
“ansible-playbook -i production site.yml -l myserver.example.com -sK” will restart nagios once.
“ansible-playbook -i production site.yml -l webservers -sK” will restart nagios once for each server in the webserver group.
Tom
Oh, I see, it's still called once for every host in that play.
That's how it currently works I'm afraid.
Slight correction –
Handlers run for each host that notified them, not for every host.
There has been some interest in a “global notify” but you may want to just do a “trailing play” that restarts the remote nagios config every time.