Hello,
I’m trying to understand these 2 processes which are interleaved with each other. I’ve been seeing this problem where checks stay into “UNKNOWN” state for > 10minutes after a config change is made on the Icinga master. As per my understanding, a config change will reload the Icinga2 process on the masters and then on the satellites. Waiting for that long is a bit making people paranoid so I’m trying to find a way to speed up this or at least find what is causing this at the first place.
Does one small config change make full config sync across the cluster or just the delta?
Probably after the reload, the load balancing occurs on the satellites. Am I right?
Do satellite try to load balance from the scratch ( full config sync) or only deal with delta coming off the config sync?
Please advise,
Thanks