Stuck on terraform init - 'Requested entity already exists'

[kenzozozo]

I'm having trouble at the terraform init step in the process.

Two issues:

1. It seems as though the process gets hung up irrevocably if it doesn't execute perfectly the first time? If the deployment.hcl file and set_env_var.sh files are not absolutely set up correctly, and the IAM groups are not set up to the exact same names, then the terraform apply step will fail, and an entirely new prefix/ENV will need to be entered in order to even attempt to run terraform again? It doesn't seem to be able to recognize when google resources are soft deleted, and I keep getting an error telling me that the "entity {PREFIX}-{ENV}-dev/folder-admin" still exists. Has this been an issue for anyone else?
2. I've been able to get further and further as I learn how to navigate the extremely touchy nature of Terraform, but could anyone confirm whether or not the IAM group names should or should not contain the domain? The reason I ask it because I got the following error: [email protected] does not exist., badRequest. I checked several times, and the group does indeed exist with the exact same spelling, but the domain is absent... Because of the above issue, however, I have to re-do the entire deployment.hcl and set_env_var.sh files, and the groups again, otherwise terraform apply will throw an error saying the entities already exist.

Any advice would be greatly appreciated.

Update: I was able to progress past this stage. The group names must contain the domain. e.g. [email protected]. This seemed to be the major hang up.

Update 2: Seems the error resurfaces during cloud run. Triggering cloud run by performing a merge in github successfully triggers the trigger, but errors out with the same "{myprefix}-{myenv}-devops already exists" error. After hunting around, it appears this is a known issue within terraform itself.

Update 3: Was able to progress past this by starting entirely from scratch: Cloning the repo, resetting group names, resetting deployment.hcl and set_env_vars.sh with all of the necessary parameters. Also deposited the requisite $10 to get my project count upped to 10 concurrent projects. Cloud build throws an error on merge:

"Apply": Error: error creating project poc9-dev-devops (poc9-dev-devops): googleapi: Error 409: Requested entity already exists, alreadyExists. If you received a 403 error, make sure you have the roles/resourcemanager.projectCreator permission

Not sure how to proceed from here.

[itsvikrams]

I'm also facing the same error.Were you able to progress further?

[kenzozozo]

Yes!

I suspect it's because the project specifies an older version of terraform, but I have no proof. Apparently, newer versions of terraform are able to delete/create/re-use projects more "intelligently," such that if the project already exists, it somehow knows to either re-use it or recreate it.

My solution was to delete all GCP projects and do everything exactly perfectly according to the directions in one shot. This means:

• deleting all projects (including removal of liens)
• re-naming security groups & config file appropriately
• deleting the entire repo and cloning it again
• following the setup instructions exactly, perfectly, in exact order

By doing this, I was able to progress to the cloud-build/git merge step, but cloud-build throws another error on merge.

If you progress past that step, please share.

[Akebed]

I'm also facing the same issue. were you able to fix it?