sorry, I"ve only just seen this, nothing to correct it's your work, I'm interested in what you're trying to do and why (more so if I can use it myself) I'm a big foreman / katello / satellite user (and it has some impressive things on it's roadmap) I"m a big fan of seeing contributions to it if it meets peoples needs rather than building something else unless that something else is a better approach or foreman doesn't solve the problem / can't solve the problem, which is why I was comparing it against what you where doing. To answer your questions • Foreman has a pretty feature complete gui - that's one of the reasons I was interested in your usecase / why you where working on this. Although the UI is currently going through some updates that I'm not a fan of, and will need time for it to settle before have a solid opinion on it • yeah, I wouldn't invest in puppet-pe or perforce products again, I understand and appreciate the desire to not use this • Foreman as an ENC is HA, but not as light as two containers, Foreman is moving to containers but still runs in a virtual machine, the virtual machine for just the ENC/web interface is pretty light and compared similar (bit larger) to the contain spec you shared in the config • foreman running just as an ENC, take around 2 minutes to deploy and sits behind a load balancer or distributed DNS name fine (I've done it behind AWS load balancers) and I've done it behind hardware load balancer and I've also done it (but not used in production) with consul advertising its components as a service (I know you use consul so thought you may find it interesting) This allows horrizontal scaling (although it never really has that much laod) and resilience, as it moves to containers this will get easier (I'm sure they would love some help containerising) • configuring foreman for your personal usecase / estate is a few hours work to get the right configuration settings and tweak them to your needs - same as any software you've never used before, however once you have the config understood for your usecase (disabled/enabled the right parameters) it takes about 2 minutes to deploy the app, I have a a clean environment on two different virtual platform that builds a host and deploys a foreman install in ~6 mins, obviously if you deploy on bare metal this is longer, the useful bit is the hammer tool and foreman API that allows post deployment configuration to be automated (which really helps with scaling and resilience) the speed of deployment really depends on what you're using (as you've said it has a lot of feautres which change the config/approach) but for context I have an AWS env that hosts just the foreman front end /ENC that's fronted with an AWS ALB and Autoscaling group, and it distributed foreman nodes across multiple AWS AZ's (and Foreman proxies too) they take about 90 seconds to fully come up and join the load balancer target group