Terraform state out of sync

[joverlee521]

Previously discussed in #10 (comment)

I'm seeing changes in the Terraform plan even though I'm starting with a fresh clone of the repo and on the main branch.

AWS_PROFILE=admin envdir ~/Repos/env.d/infra-github-token/ terraform -chdir=env/production init
AWS_PROFILE=admin envdir ~/Repos/env.d/infra-github-token/ terraform -chdir=env/production plan -out=plan
...
Plan: 38 to add, 0 to change, 4 to destroy.

plan.txt

[victorlin]

I can't read the file linked in the issue description. Here's my output with terraform plan -no-color:

plan.txt

[joverlee521]

@victorlin Weird, I can't read your file either

$ terraform show ~/Downloads/plan.txt
╷
│ Error: Failed to read the given file as a state or plan file
│ 
│ State read error: Error reading /Users/joverlee/Downloads/plan.txt as a statefile: 2 problems:
│ 
│ - Unsupported state file format: The state file could not be parsed as JSON: syntax error at byte offset 1.
│ - Unsupported state file format: The state file does not have a "version" attribute, which is required to identify the format version.
│ 
│ Plan read error: couldn't load the provided path as either a local plan file (zip: not a valid zip file) or a saved cloud plan (invalid
│ character 'd' looking for beginning of value)

[victorlin]

Oh... I didn't realize plan.txt referred to the actual plan file used for terraform show, I thought it was just the full output. Now we have both – mine is the console output of terraform plan and yours is the plan file.

[joverlee521]

Terraform reading...

The "add" plans are resources that rely on Terraform state:

When Terraform creates a new infrastructure object represented by a resource block, the identifier for that real object is saved in Terraform's state, allowing it to be updated and destroyed in response to future changes.

Terraform state:

This state is stored by default in a local file named "terraform.tfstate", but we recommend storing it in HCP Terraform to version, encrypt, and securely share it with your team.

We are currently excluding the local state file from version control in the .gitignore

OH, but we do have a backend defined in terraform.tf

[joverlee521]

$ AWS_PROFILE=admin aws s3 ls s3://nextstrain-terraform/infra/production/tfstate
2024-04-08 17:28:45      16127 tfstate

The remote tfstate file was last updated in April...which doesn't sound right considering all of the activity in this repo over the past month.

Ah, but there's been a lot of updates to s3://nextstrain-terraform/workspace/trs/infra/production/tfstate.

@tsibley, are you working from a temporary workspace?

[tsibley]

Ugh, I am.

$ envd aws/blab/trs-admin-read-only github/tsibley/nextstrain-admin -- terraform -chdir=env/production workspace list
  default
* trs

Dammit. I'll move the state over (tonight or tomorrow morning) and ping here when I do.

[tsibley]

Moved with envd aws/blab/trs github/tsibley/nextstrain-admin -- bash -c 'cd env/production; terraform state pull | (terraform workspace select default && terraform state push -)'.

@joverlee521 @victorlin What does plan say for you now?

[joverlee521]

No changes. Your infrastructure matches the configuration.

Terraform has compared your real infrastructure against your configuration and found no differences, so no changes are needed.

🎉

[tsibley]

Excellent. Sorry for the hiccups!

[tsibley]

(Need to get the current Terraform workspace into my prompt, since "try not to use them much and be careful" hasn't been enough…)