4.1 KiB
Nanobox Guide
Nanobox lets you develop apps in an environment identical to (or at least nearly so) the environment it will deploy to. It supports deploying to multiple cloud providers, so you have lots of choices about where your instance will run.
Development
You will need Nanobox installed, along with Docker if you're on Linux (Windows
and macOS can use Docker Native, but the bundled VirtualBox is more performant
while the Docker team works out some filesystem speed issues). Once Nanobox is
installed, the process is simple - clone the Mastodon repo, set a few variables
with nanobox evar add local {VARIABLE}={value}
(see below on which ones need
to be set), and run nanobox run
to get to a console. It will take some time to
build your local dev environment, but once it's done, simply set up the DB using
bundle exec rake db:setup
as normal, and you're off.
Production
To deploy, you'll need to create an application in your Nanobox
dashboard (which requires a free
Nanobox.io account), clone the
Mastodon repo (if you haven't already set up the local development environment),
set the new app as your deploy target with nanobox remote add {app-name}
, set
up the variables below using either nanobox evar add {VARIABLE}={value}
or the
app's dashboard, and then run nanobox deploy
. It will take a while to build
everything for deployment the first time, but the process handles itself.
Updating
To update in production, simply grab the latest tagged version with git fetch && git checkout $(git tag -l | grep -v 'rc[0-9]*$' | sort -V | tail -n 1)
, then
re-run nanobox deploy
- Nanobox automatically handles the rest. Updates should
be much faster than the initial deploy, since the build should be cached on your
own computer, and only changes are sent to the server itself. Either way, you
shouldn't see any downtime while the update process runs.
For development, grab the latest tagged version using the instructions above,
then use a nanobox run
console to perform all the upgrade steps in the Release
Notes (except for installing dependencies - the boxfile.yml
is updated as
needed to pull those in automatically).
Environment Variables
Mastodon will not run under Nanobox without first setting a handful of required variables:
-
RAILS_ENV
- set this toproduction
, unless you're in development (this value is treated asdevelopment
if it isn't set) -
NODE_ENV
- same asRAILS_ENV
-
PAPERCLIP_SECRET
- set to a random string of characters; you can usenanobox run bundle exec rake secret
to generate one -
SECRET_KEY_BASE
- set to a random string of characters; you can usenanobox run bundle exec rake secret
to generate one -
OTP_SECRET
- set to a random string of characters; you can usenanobox run bundle exec rake secret
to generate one -
VAPID_PUBLIC_KEY
- generated by runningnanobox run bundle exec rake mastodon:webpush:generate_vapid_key
-
VAPID_PRIVATE_KEY
- generated at the same time asVAPID_PUBLIC_KEY
You can also set some optional values, which should override the defaults in
.env.nanobox
:
-
LOCAL_DOMAIN
- set to whatever domain you want to use as your instance name; defaults to{app-name}.nanoapp.io
, which is provided by Nanobox for use as a CNAME target -
SINGLE_USER_MODE
- set this totrue
if you want to run a single-user instance; default is unset
And really any other setting you'd normally put into .env.production
, such as:
-
SMTP_SERVER
- your SMTP server's address; default is blank -
SMTP_PORT
- your SMTP server's port number; defaults to 587, which is almost always correct -
SMTP_LOGIN
- your SMTP username; default is blank -
SMTP_PASSWORD
- your SMTP password; default is blank -
SMTP_FROM_ADDRESS
- this instance's emails will come from this address; defaults tonotifications@{app-name}.nanoapp.io
-
etc...