Issue: delete_automated_backups = false Not Preventing Backup Deletion During RDS Instance Replacement for Point-in-Time Recover

Terraform
1

I am encountering an issue with using the delete_automated_backups = false argument in the AWS RDS module while trying to perform a point-in-time recovery (PITR) on an RDS instance.

Here’s the scenario:

  • I am using Terraform to manage an RDS instance and I wish to restore it using a specific point in time.
  • The process requires passing the RDS instance name and the desired recovery time.
  • However, when I apply the Terraform configuration, the RDS instance is being deleted and then recreated, which, as per Terraform documentation, should not delete the automated backups when delete_automated_backups = false is set.
  • Despite setting this argument to false, the backups are still being deleted during the apply phase. This causes issues because in order for PITR to work, the automated backups should remain available, but they are deleted as part of the deletion and recreation of the RDS instance.

Expected Behavior:

  • The delete_automated_backups = false argument should prevent the deletion of automated backups during the replacement of the RDS instance, ensuring that backups are available for point-in-time recovery.

Actual Behavior:

  • The automated backups are deleted, which interferes with performing point-in-time recovery on the new instance after Terraform applies the changes.