AWS ALB takes a very long time to provision

Terraform Version


AWS Provider: 2.59.0

Expected Behavior

Spinning up an amazon application load balancer should take no more than 5 minutes

Actual Behavior

Amazon Load Balancers can sometimes take up to 30 minutes to deploy

Steps to Reproduce

We’ve noticed an issue starting with 2.58 and higher (Can’t remember if we’ve had it on older versions to be honest) where spinning up an AWS ALB can take upwards of 30 minutes. On most of our ci/cd runs it takes no more than 5. We’ve recently turned on DEBUG to trace and have the logs however I can’t most it here since there could be potentially sensitive information in those files. Would anyone happen to know the best way to parse this file and see if this is a terraform or aws api issue?

Our load balancer gets applied through a module which has the following resources:

  1. data.aws_acm_certificate
  2. data_aws_subnet_ids
  3. resource.aws_lb
  4. resource.aws_lb_target_group
  5. resource.aws_lb_target_group_attachment
  6. resource.aws_lb_listener

Now I realize with that many resources in the module it is a bit hard to pinpoint exactly what is causing the issue, but I’m hoping either someone has faced the same thing where ALBs can take a long time to provision with a fix or help me in parsing through the debug log to see exactly where the issue is. Thank you so much!