Switching HTTPS Termination to the AWS Load Balancer

[FinnIckler]

This change needs a redeployment of the Server!

Changes done:

• Removing the acme.sh cron jobs to renew the certificate
• Removing the Certificate Check from the server controller
• Changing the base.rb Chef files to not create the https nginx config
• Changing the nginx config to server traffic on port 80
• Changing the servers.sh to
  • SSH into the already existing server via elastic ip not hostname (as the hostname will be attached to the load balancer)
  • Doing the health check using the load balancer internal health check
  • Register the targets with the load balancer in the passthetorch phase
  • Deregistering the targets in the reap phase
• Changing bootstrap.sh to not sync the certificates

This also includes infrastructure changes

• Creating a Load Balancer with a *.worldcubeassociation.org certificate
• Creating a target group for staging and production

While I have tested the code/infrastructure changes on staging, I have not made the necessary DNS changes for production. So my recommendation is, after this is reviewed, we start a new staging and then prod server with the changes, but we do not reap the server. Instead, we perform a DNS fallover to the loadbalancer after checking everything works as intended to achieve a no downtime migration.

PS: Sorry for all the commits, I had to rebase a commit out of there

[gregorbg]

The commit history is a hot mess. Please fix this before proceeding with the review.

[FinnIckler]

I am just going to close and reopen this one