Image to setup wordpress in production with versioned plugins
wordpress:wordpress
Checking the plugins :
- Go to the admin interface and note down the versions of each new plugin
- Update them
- Run
docker-compose up - Check if everything works
- if it does, update the DockerFile (https://github.com/CoderDojo/wordpress-docker/blob/master/Dockerfile)
- When merged, the image will be published
Both instances lives on ECS. ECS requires at least 50% health, you need 2 instances at startup. However, plugins run their own migrations, on a shared DB.
- scale down to 1 instance
- do the plugins update on the live instance (db migrations)
- wait for the new image to be published
- deploy on ECS
Currently living on ECS:
- Under "Task Definitions" create a new revision of the task called "wordpress-fargate" (
Task Definitions > wordpress-fargate > xx) once the Docker image is published to DockerHub (this should bump up the auto-incrementing revision number eg.wordpress-fargate:xx) - Edit the service in the Cluster "wordpress-single-cluster" (open the service, hit "Update" top right)
- Change the revision number to the latest task revision
- Next until the end
- The service (Clusters -> wordpress-single-cluster -> Tasks) should show a new task having been started / provisioned (with the task definition showing the appopriate revision number at the end).
If it doesn't:
- Edit the service and go to the Autoscaling tab, hit "Update" and then skip ahead to "Service Auto Scaling (optional)"
- Change Minimum no. of tasks and Desired no. of tasks to
2 - Head back to the "Tasks" tab of the cluster, and now the new task should appear (as above).
- Once provisioned / running, scale the cluster back down to
1, head back to the "Tasks" tab and stop the task with the outdated task definition
- sucuri: Verify if confs/sucuri-settings changed Verify the .htaccess-es
- CDF role (blacklist of capabilities):
Add to the blacklist and deploy or Remove from the blacklist, deploy and update in admin-panel