we start with Icinga2 and I have general questions about this.
Should we start with the director module or should we add director later?
Our target is to reach a basic monitoring, later we want to expand to a cluster and high availability.
It would be unlucky if we made the wrong decisions by defining the project scope.
Are there any best practice recommendations?
When you intend to use the director then you should do this right from the beginning as there is (currently) no feature to import your host and service definitions.
Hi and welcome,
my experience is to decide for one way. If you built your config via command line and add the director your config will crash or becomes chaos. I dont use the director and it’s the best way for me.
I would recommend taking the time to evaluate all possible methods, best with a test VM. You may use the Icinga Vagrant boxes for this purpose too.
Generally speaking it helps to understand the configuration DSL even when working with the Director or any sort of config management tool. For example, when the deployment fails with an error from the core’s config validation. Or when building a distributed environment where zones and endpoints still need to be managed on each node.