Skip to content

SeAT

General Upgrades

As with anything, it is a very good idea to have backups ready before attempting any upgrades. In the case of SeAT, the most important component needs to be backed up is the SeAT database. Should something go wrong, then you can simply re-install SeAT, restore the database, and you should be good to go.

Docker installation

As expected, updates for SeAT are deployed via Dockerhub and the images within the eveseat organization. Every package version release will automatically start the build process to generate a new docker image. This means updates are super simple in the docker world. To update your instance, simply run:

# Update to the latest dockerhub images
docker-compose pull

# Apply the updates to your instllation
docker-compose up -d

# Cleanup any dangling images
docker image prune -f

Better safe then sorry

Always perform a database backup of your database before doing an update. Always.

Blade installation

  • Ensure that you are in the path where you installed. By default this should be /var/www/seat.
  • Cut jobs processing.
supervisorctl stop all
  • Put your application into maintenance mode. This will ensure that no request from the outside will hit your applications logic, and also help you perform an upgrade uninterrupted. Do this with:
php artisan down
  • Its a good thing to update composer itself. This is not a hard requirement, but definitely recommended. Do this with:
composer self-update
  • The next thing to do is get the latest SeAT packages. The below example adds the --no-dev argument as these packages are generally not needed in production. Upgrade the packages with:
composer update --no-dev
  • With the packages upgraded, we can now publish any new migrations/assets/configs. This can be done with:
php artisan vendor:publish --force --all
  • Once the needed files are published, run any new/outstanding migrations with:
php artisan migrate
  • With the migrations done, run the seeders to update any static data:
php artisan db:seed --class=Seat\\Console\\database\\seeds\\ScheduleSeeder
  • Finally, we can simply bring our application back out of maintenance mode with:
php artisan up
  • Prune job queue, this will prevent already queued jobs to use old codebase
php artisan cache:clear
  • Renew cached configuration and routes
php artisan config:cache
php artisan route:cache
  • Restart the supervisor workers to ensure they also pickup the latest code:
supervisorctl start all

Better safe then sorry

Always perform a database backup of your database before doing an update. Always.

You can use mysqldump -uroot -p seat > backup.sql (change root and seat according to your configuration)