Re: [Hawkular-dev] About configuration

Honestly, once I am using a configuration management system (like Puppet) it doesn't really matter that it is another file, especially once someone has released a puppet module to do all the lifting. I recently deployed (on a side project) collectd, graphite, grafana, and the ELK stack without touching any of their config files directly and specifying only a minimal set of values that differed from convention.   If the goal is to work well with tooling used by sys admins, like Puppet/Chef/etc, then we really should develop and release our own modules for whatever we want to support. If configuration is file based, this will take the developer maybe a couple hours for each supported toolset.   If the configuration is inside EAP, then it is still possible, just difficult. Don't expect the community to do it on their own. It requires integrating non-java languages with the management interfaces. You can see an example of what is required for doing this with Puppet in the module I linked before (and it can be used for configuring hawkular if its config is in container). Similar work would need to be done for whatever other config management chosen. Expect to spend several weeks for each variant going this way. ----- Original Message -----