terraform recreating exact same ressources (in docker context with persisted volume)

Question

I have a basic understanding but am only 2 days into my terraform experience, hence my problem.

I have a grafana container with persisted data.

Terraform runs on a different container and is executed on deployment. Hence, each time I wish to deploy to an existing grafana volume I encounter terraform errors "already exists" or "managed by someone else".

I would like to "replace" or "overwrite" any existing, regardless of any changes made in between but -replace="<existing.resource>" does not appear to address the "existing resource" problem.

Destroy does not work as I have not imported anything, I have yet to try to import into a skeleton resource, destroy and apply and I would like not to.

What am I missing? Does my approach comply with the declarative approach of terraform? How can I "remind" terraform about the existing resources without to much condition?

The problem appears to be a lack of understanding.

The main.tf creates a datasource and an alert. Dashboards are created using grafonnet, they are not relevant.

terraform {
   required_providers {
      grafana = {
         source  = "grafana/grafana"
         version = "1.30.0"
      }
   }
}

provider "grafana" {
   url   = ...
   auth  = ...
}

resource "grafana_data_source" "influxdb" {
  uid = "datasource_influxdb"
  type = "influxdb"
  name = "influxdb"
  url  = "..."
  json_data_encoded = jsonencode({
    defaultBucket     = "..."
    httpMode          = "POST"
    organization      = "..."
    timeout           = 5
    tlsAuth           = false
    tlsAuthWithCACert = false
    version           = "Flux"
    tlsSkipVerify     = true
  })
  secure_json_data_encoded = jsonencode({
    token = "..."
  })
}

 
resource "grafana_folder" "rule_folder" {
    title = "Alerts"
    uid = "rule_folder_uid"
}

resource "grafana_rule_group" "my_alert_rule" {
    name = "My Rule Group"
    folder_uid = grafana_folder.rule_folder.uid
    interval_seconds = 240
    org_id = 1
    rule {
        name = "Alerted CPU"
        for = "2m"
        condition = "B"
        no_data_state = "NoData"
        exec_err_state = "Alerting"
        annotations = {
            __dashboardUid__ = "..."
            __panelId__ = 3
        }
        ...
   }
}

The entrypoint.sh applies terraform using the official terraform image (which curiously lacks an actual terraform installation):

#!/bin/bash


wget https://releases.hashicorp.com/terraform/1.3.4/terraform_1.3.4_linux_amd64.zip
unzip terraform_1.3.4_linux_amd64.zip
mv terraform /usr/bin/terraform

terraform init 
terraform apply -auto-approve 

terraform apply -auto-approve -replace="grafana_data_source.influxdb"  #problem: terraform does not control this resource

tail -f /dev/null

The log-feed is as follows:

Plan: 3 to add, 0 to change, 0 to destroy.

grafana_folder.rule_folder: Creating...

grafana_data_source.influxdb: Creating...


│ Error: status: 409, body: {"message":"data source with the same name already exists","traceID":""}

│   with grafana_data_source.influxdb,

│   on main.tf line 15, in resource "grafana_data_source" "influxdb":

│   15: resource "grafana_data_source" "influxdb" {

│ Error: status: 412, body: {"message":"the folder has been changed by someone else","status":"version-mismatch"}

│   with grafana_folder.rule_folder,

│   on main.tf line 36, in resource "grafana_folder" "rule_folder":

│   36: resource "grafana_folder" "rule_folder" {

It would be great to see some code, so please add it to the question. — Marko E, Nov 14 '22 at 18:58
So the issue is with the resource that already exists? If it was created outside of terraform, you can import it to your state file. — Marko E, Nov 15 '22 at 15:46
I may have understood; the missing link was understanding the ID of the desired resource: This was helpful: https://stackoverflow.com/questions/65678433/terraform-import-an-existing-aws-resource-to-a-module — Christoph, Nov 15 '22 at 15:58

score 0 · Accepted Answer · answered Nov 15 '22 at 16:01

The answer was the addition of explicit imports. Beware to post the correct ID: terraform import an existing AWS resource to a module

In my case

terraform  import grafana_data_source.influxdb datasource_influxdb
terraform  import grafana_folder.rule_folder rule_folder_uid

terraform recreating exact same ressources (in docker context with persisted volume)

1 Answers1