Re: [atomic-devel] Structure of Atomic Host docs for Fedora and CentOS

On Wed, Dec 7, 2016, at 04:22 AM, Jason Brooks wrote:

On Dec 6, 2016 17:22, "Josh Berkus" <jberkus redhat com> wrote:

Folks,

Given all of the changes in the platform, it's time to rewrite and

re-organize the Atomic Host documentation. Here's what I'm thinking for

structure:

* Introduction

* Installation

* Vagrant

* Public Cloud

* ISO

* Others

* Quick Start

* Links to simplest setups, below

* Cluster Setup

* Single-Node Vagrant Example (QS)

* AH + Kubernetes

* Kubeadm (QS)

* Kubernetes-Ansible (production)

* OpenShift

* oc cluster up (QS)

* Openshift-Ansible (production)

* Other Platforms

* Swarm?

* Mesos?

* Using Ostree

* Updates

* Rebasing

* Hotfixes & Testing

* Overlays

* Atomic Tools

* Atomic CLI (link, maybe)

* Docker (and Docker Latest)

* System Containers

* Other stuff?

Open Questions:

1) Should these be combined Fedora/CentOS docs, or should we make a

"build" for each OS?

I think they should work for both, be hosted from projectatomic or rtd, and then excerpted or linked to as desired by Fedora and CentOS.

I believe we should host them where it makes the most sense for Project Atomic. I am partial to projectatomic.io, but can easily see using RTD as a temporary stop-gap.

I hope the platform we are building in Fedora will be able to consume the source files for rebranding and publishing as is appropriate to Fedora (+RHEL +CentOS). Once finished (heh) that platform would allow for easy hosting at projectatomic.io as well.

2) Where should these live? projectatomic.io or RTD?

3) Is Vagrant really where we want to direct people as a quickstart

option? I'm asking because I'm not that clear on what users can *do*

with an single atomic node on vagrant.

A single node can get you a taste of docker or origin or kube, esp if you're on Windows or OS X. You could develop for a legit cluster using a single node vm.

--
--
Josh Berkus
Project Atomic
Red Hat OSAS