Clearer Documentation for beginners

[dfreilich]

Is your feature request related to a problem? Please describe.
My partners and I attempted to set up CFCR during our Onboarding Week, but weren't able to successfully do it. The directions on how to set it up here were outdated, and the README referred to DNS requirements, without explaining how to set those up, or providing a terraform.

Describe the solution you'd like
Clearer documentation, with links/steps on how to create those deployment requirements (DNS-NAME/LOADBALANCER_ADDRESS), or a terraform template to run, would have helped tremendously in setting it up.

Describe alternatives you've considered
Additionally, updating the https://docs-cfcr.cfapps.io/ site with updated step-by-step documentation would have helped a lot.

Additional context
Add any other context about the feature request here.

[pydctw]

Thanks for the input. As you mentioned, this blog is not an official documentation so it is not maintained and will not reflect the changes in CFCR.

You can follow this recommended procedure to setup BOSH on GCP and deploying CFCF instruction.

We will give more info on DNS-NAME/LOADBALANCER_ADDRESS as requested.

[aegershman]

Not to pile on but it may be beneficial for more top-level examples of leveraging bbl for cfcr. At the moment it appears bbl has examples for standing up an env for Concourse or CF, but mentions of cfcr are under the advanced-configuration and implemented as a plan patch. So it's a little unclear whether bbl should be used with a plan patch from the bbl repo, or if bbl should be ran as bbl up without any extra config to get a plan ol' director && have the rest of the deployment be handled with opsfiles from kubo-deployment... Or if there needs to be a terraforming process outside of anything related to BOSH?

Personally I feel it would be helpful for more streamlining of terraform for setting up LBs, DNS, etc. (or maybe some examples?), but perhaps this is the responsibility of bbl plan patches.

With a more streamlined approach (through doc and/or as part of bbl itself) this would be the easiest way to get manageable k8s clusters up without having to use k8s-specific CLI tools like kops, eksctl, etc., which don't offer the best cluster-lifecycle experience.

Apologies, don't mean to hijack or pontificate, just wanted to add more thoughts on it.