feat: add identifier prefixes to recreate resources

[nitrocode]

what

• feat: add cluster identifier prefixes

why

• We can avoid name conflicts upon recreation by setting the following inputs
  • parameter_group_name_prefix_enabled
  • rds_cluster_identifier_prefix_enabled
  • db_subnet_group_prefix_enabled
• One downside is that the *_prefix arguments add an additional 27 characters to each resource which can cause errors upon apply due to it exceeding 64 characters
  • For this reason, it may be advisable to NOT use *_prefix and instead use a random string for each of these resource types

references

• Closes Cannot recreate database instance due to create_before_destroy set to true on the cluster instance #257
• Related to @finchr's PR reopen #192 #213
• https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/rds_cluster#cluster_identifier_prefix-1
• https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/security_group.html#name_prefix-1

[nitrocode]

/terratest

[finchr]

The unique name for the aws_rds_cluster_instance should be controlled by the random_pet that feeds into the module.rds_identifier. Why is that not working here?

[nitrocode]

How do you know it's not working?

[nitrocode]

Oh I see what you're saying. The random pet does go into the rds_identifier. Hmm, maybe I'm not using it correctly

[finchr]

@nitrocode there is some history here, I made some changes in #236 but after some FUD some of the changes were taken out in #242. You would probably need to add cluster_identifier back into the random_pet to get the features you want.

[mergify]

💥 This pull request now has conflicts. Could you fix it @nitrocode? 🙏