I’m in the process of bringing up an over year old installation of Trellis (using the steps found in this guide: http://www.chrisknightindustries.com/2015/24/11/git-subtrees-for-trellis-workflow.html) from pre-1.0 to 1.2.0 — which has since added Ubuntu 18.04 support and started defaulting to it. The initial install was running a different version of PHP(7.2) and Ubuntu 16.04 Xenial LTS.
I am wondering if re-provisioning the server with Ansible will bring up and update Ubuntu, or if I will need to spin up a separate droplet with a clean installation of 18.04 instead? Just curious about the best practices here to ensure there are no problems with the production environment. Any advice or help would be greatly appreciated.
I’d recommend just spinning up a new droplet and this way, it allows you to also setup a temporary staging environment. You can install any plugin updates and flip the DNS over when you are comfortable.
Would a Trellis upgrade process be within the scope of the Trellis docs? I know upgrading Trellis projects kind of depends on the way you have your project set up and your project requirements, but it took quite a bit of fumbling and confusion through discourse before we had an upgrade procedure set because it’s not documented. The recommendation of using a Digital Ocean floating IP to allow no-downtime switches between environments was also a life saver in our case, as we service websites for customers who manage their own DNS and, how do I put this, don’t always have their finger on the pulse of their domain management logins.
I’d be happy to draft something on our setup based on @fullyint’s process here if that’s helpful, but I’m not 100% sure that we’re following upgrade best practice (e.g. we use Trellis as a git subtree in our projects to simplify the upgrade / merge process).