Upgrading a Stack to a Later Terraform Version

Upgrade a stack in Resource Manager to a later Terraform version.

Note

These instructions don't apply to Resource Manager stacks created through Marketplace.

These steps are completed in the command line and the Console.

For information about Terraform versions supported by Resource Manager, see Supported Terraform Versions.

Before You Begin

To successfully upgrade your stack, you must have the following requirements:

A MacOS, Linux, or Windows computer for running command line tools
Software to create and unpack .zip archives, such as 7-ZIP
IAM policies to manage stacks and jobs

Automatic Upgrade

To support deprecation of older Terraform versions (earlier than 1.5.x), Resource Manager will try an automatic Terraform version upgrade. This automatic upgrade will be attempted whenever you run a successful apply job (job state is succeeded) on a stack that's configured for an older Terraform version.

Note

The target version for the attempted upgrade depends on the stack's currently configured Terraform version. Intermediary automatic upgrades move the stack toward 1.5.x in one or more successful upgrade attempts. Using intermediary upgrades works around differences in HCL syntax, provider block definitions, and state file differences, as discussed in the manual upgrade instructions. The target version for an upgrade is selected as follows:

The target version to upgrade 0.12.x is 0.13.x.
The target version to upgrade 0.13.x is 0.14.x.
The target version to upgrade 0.14.x, 1.0.x, 1.1.x, or 1.2.x is 1.5.x.

The upgrade attempt only occurs after an apply job succeeds (job state is succeeded).

If the upgrade succeeds, then the stack's Terraform version and state file are upgraded to 1.5.x (or intermediary version). No further action is required.
If the upgrade fails, then the stack's Terraform version is unchanged, and resources aren't affected. An upgrade failure message appears in the stack's details page. In this situation, get logs for the apply job (select Show terraform upgrade logs for information about the failure) and then update the Terraform configuration. The next successful apply job that you run on this stack will trigger the automatic upgrade process by Resource Manager.
Note

Even if the upgrade fails, the job state is still succeeded.

Upgrade Paths

Following are supported upgrade paths by initial version.

Initial Terraform Version	Upgrade Path
0.12	In sequence, upgrade to each supported version: 0.13, 0.14, 1.0 (recommended), 1.1, 1.2, 1.5.
0.13	In sequence, upgrade to each supported version: 0.14, 1.0 (recommended), 1.1, 1.2, 1.5.
0.14	In sequence, upgrade to each supported version: 1.0 (recommended), 1.1, 1.2, 1.5.
1.0	In sequence, upgrade to each supported version: 1.1, 1.2, 1.5.
1.1	In sequence, upgrade to each supported version: 1.2, 1.5.
1.2	Upgrade to 1.5.

See Terraform's official upgrade guides:

Task 1: Confirm Up-to-Date Infrastructure

This task uses the Console. For CLI and API instructions for a step, see the associated link.

On the Stacks list page, select the stack that you want to work with. If you need help finding the list page or the stack, see Listing Stacks.
Confirm that the details page for the stack shows the expected version (Terraform version).
For CLI and API instructions to get a stack's details, see Getting a Stack's Details.
Check for pending changes to infrastructure:
1. Select Plan.
  
  When the action finishes, the details page for the related job opens.
  
  For CLI and API instructions, see Creating a Plan Job.
2. To review the log for the completed plan action, select Logs.
  
  The log contents indicate whether the stack is up-to-date or has pending changes.
  
  For CLI and API instructions, see Getting Logs for a Job.
  
  Example for a stack that's up-to-date (no pending changes):
  
  No changes. Infrastructure is up-to-date.
If the log contents indicate pending changes, then apply pending changes:
1. Select Stack details to go back to the details page for the stack.
2. Select Apply.
  
  For CLI and API instructions, see Creating an Apply Job.
  
  When the action finishes, the details page for the related job opens.
3. Confirm that the apply action was successful: Select Logs to review the log for the completed plan action.
  
  For CLI and API instructions, see Getting Logs for a Job.

When the stack infrastructure is up-to-date, you can proceed to the next task for downloading Terraform configuration and state files.

Task 2: Download the Configuration and State

Note

If the stack's Terraform configuration is stored in a source code control system, such as GitLab, then check out and download the Terraform configuration from there.

If the stack's Terraform configuration is stored in a bucket, then download the Terraform configuration from there.

This task uses the Console. For CLI and API instructions for a step, see the associated link.

On a computer that can run command line tools, create a folder to store the downloaded Terraform configuration and state.
Example folder name: c:\StackUpgrade
In the Console: On the Stacks list page, select the stack that you want to work with. If you need help finding the list page or the stack, see Listing Stacks.
Next to Terraform configuration, select Download.

For CLI and API instructions, see Getting a Stack's Terraform Configuration.
Download the Terraform state file:
1. On the stack's details page, select View state.
2. To download the stack's state file, go to More actions and select Download Terraform state.
For CLI and API instructions, see Getting a Stack's State File.

When the Terraform configuration and state are downloaded, you can proceed to the next task for upgrading the Terraform configuration.

Task 3: Upgrade the Configuration

This task provides customized steps for upgrading a Terraform configuration used with Resource Manager.

On the same computer that you used to store the downloaded Terraform configuration and state files, download the .zip files needed to upgrade your Terraform configuration:
- From https://releases.hashicorp.com/terraform, download the .zip file for your target Terraform version and operating system.
  
  For example, for Terraform 0.13 on Mac, download the Darwin version: https://releases.hashicorp.com/terraform/0.13.7/terraform_0.13.7_darwin_amd64.zip.
Extract the contents of each .zip file.

Update the provider configuration to add arguments such as user_ocid, fingerprint and private_key_path. You might have commented out these arguments before.

Example of commented-out arguments:

provider "oci" {
                    tenancy_ocid = var.tenancy_ocid
                    /*
                    user_ocid = var.user_ocid
                    fingerprint = var.fingerprint
                    private_key_path = var.private_key_path
                    */
                    region = var.region
                    }

To follow code examples in the rest of this procedure, rename the extracted file as terraform_<major-version>.
Example: terraform_13
To make the command (extracted file) accessible, store it in a directory that's present in your PATH directory.
Open a command prompt.
Change directory to the folder where you stored the downloaded stack information.
Example:
```
cd c:\StackUpgrade
```
To initialize Terraform, run the following command:
```
terraform_<major-version> init
```
Example:
```
terraform_13 init
```

To upgrade the syntax of your Terraform configuration, run the command for the target Terraform version:

Target Terraform Version	Command
0.13	`terraform_13 13upgrade`
0.14	`terraform_14 14upgrade`
1.0 and later	None needed. If preceding upgrades were successfully applied during the process, then no special changes are needed for upgrading the configuration.

The output indicates whether the upgrade was successful.

If successful, continue with the next step.

If not successful, then make manual changes to your Terraform configuration files as directed.

Create a .zip archive of your Terraform configuration files.

Example .zip archive: c:\StackUpgrade\MyConfigUpgraded.zip

Ensure that the archive omits the Terraform state file (terraform.tfstate) and the .terraform directory to satisfy the required file structure for Terraform configurations.
If a source code control system (such as GitHub) is used for the stack's Terraform configuration, then commit the upgraded Terraform configuration there.
The most recent commit is used when you run jobs on the stack.
If an Object Storage bucket is used for the stack's Terraform configuration, then change the contents of that bucket to correspond with the upgraded Terraform configuration.

Note

Back up the current bucket before changing it to correspond to the upgraded Terraform configuration. Limit the bucket to files that are intended for use with Terraform.

The most recent contents of the bucket are used when you run jobs on the stack.

When the Terraform configuration is successfully upgraded to the target Terraform version, then you can proceed to the next task for upgrading the stack.

Task 4: Upgrade the Stack

This task uses the Console. For CLI and API instructions for updating a stack, see Updating a Stack.

In the Console, reopen the details page for the stack that you're upgrading: On the Stacks list page, select the stack that you want to work with. If you need help finding the list page or the stack, see Listing Stacks.
From the Actions menu for the stack, select Edit.
Upload the upgraded Terraform configuration to the stack.

You can either drag the file onto the dialog's control or select Browse and navigate to the location of the file or folder.

Example file path: c:\StackUpgrade\MyConfigUpgraded.zip

The dialog box is populated with information contained in the Terraform configuration.

Note

Skip this uploading step if the stack's Terraform configuration is stored in a source code control system (such as GitHub) or in an Object Storage bucket, then the stack was configured to use the upgraded Terraform configuration in Task 3: Upgrade the Configuration when you committed the change to source code or uploaded the file to the bucket.)
Specify the target Terraform version: Change the Terraform version.
Select Next twice, then select Save changes.

The stack is now synchronized with the upgraded Terraform configuration and specified Terraform version. You can now proceed to the next task for importing the state file.

Task 5: Import the State

This task uses the Console. For CLI and API instructions, see Creating an Import Job.

In the Console, reopen the details page for the stack that you're upgrading: On the Stacks list page, select the stack that you want to work with. If you need help finding the list page or the stack, see Listing Stacks.
Go to More actions and select Import state.
In the Import panel, add the downloaded Terraform state file.

You can either drag the file onto the dialog's control or select Browse and navigate to the location of the file or folder.

Example file path: c:\StackUpgrade\terraform.tfstate
Select Import.

The import job is created. The new job is listed under Jobs.

When the job finishes, the Job details page opens.
Confirm successful import: Select Logs.

For CLI and API instructions, see Getting Logs for a Job.

After a successful import of the state file, proceed to the next task for checking for issues.

Task 6: Check for Issues

This task uses the Console. For CLI and API instructions for a step, see the associated link.

On the Stacks list page, select the stack that you want to work with. If you need help finding the list page or the stack, see Listing Stacks.
Check for pending changes to infrastructure:
1. Select Plan.
  
  When the action finishes, the details page for the related job opens.
  
  For CLI and API instructions, see Creating a Plan Job.
2. To review the log for the completed plan action, select Logs.
  
  For CLI and API instructions, see Getting Logs for a Job.
3. In the log contents, check for issues described at Troubleshooting Logs During an Upgrade.
Resolve any listed issues by manually updating the Terraform configuration, as described at Troubleshooting Logs During an Upgrade.
On the details page for the stack, select Edit, then upload the newly updated Terraform configuration to the stack.

You can either drag the file onto the dialog's control or select Browse and navigate to the location of the file or folder.

Example file path: c:\StackUpgrade\MyConfigUpgraded.zip

The dialog box is populated with information contained in the Terraform configuration.

Note

Skip this uploading step if the stack's Terraform configuration is stored in a source code control system (such as GitHub) or in an Object Storage bucket, then the stack was configured to use the upgraded Terraform configuration in Task 3: Upgrade the Configuration when you committed the change to source code or uploaded the file to the bucket.)
Select Next twice, then select Save changes.
Run the plan action again to confirm that the issues are no longer listed in the associated log contents.
1. Select Plan.
  
  When the action finishes, the details page for the related job opens.
  
  For CLI and API instructions, see Creating a Plan Job.
2. To review the log for the completed plan action, select Logs.
  
  For CLI and API instructions, see Getting Logs for a Job.

Troubleshooting Logs During an Upgrade

Following are some issues you might see in logs while upgrading a stack to a new Terraform version. This list is limited to a few issues that the Resource Manager service team knows about.

Error: Failed to install providers

The log shows an error message similar to the following.

Error:  Failed to install providers
Could not find required providers, but found possible alternatives:
hashicorp/gitlab -> gitlabhq/gitlab
If these suggestions look correct, upgrade your configuration with the following command:
terraform 0.13upgrade .

The configuration doesn't meet requirements for the specified Terraform version. Version 0.13.x and later don't use this syntax for providers. Example configuration causing this error:

provider "gitlab" {
  token = "glpat-_abcd"
}  
# Add a project owned by the user
resource "gitlab_project" "sample_project" {
  name = "example"
}

Add a required_providers block and explicitly mention the source information for the provider. For more information, see Requiring Providers. Example update:

terraform {
  required_providers {
    oci = {
      source  = "oracle/oci"
      version = "5.46"
    }
    gitlab = {
      source  = "gitlabhq/gitlab"
      version = "17.8.0"
    }
  }
}

Error: Invalid quoted type constraints

The log shows an error message similar to the following.

Error: Invalid quoted type constraints on variables.tf line 18, in variable "vcn_dns_label"
 18:       type = "string"
Terraform 0.11 and earlier required type constraints to be given in quotes, 
but that form is now deprecated and will be removed in a future version of
Terraform. Remove the quotes around "string".

The configuration doesn't meet requirements for the specified Terraform version. Version 1.0.x and later don't use quotes for type declarations of variables. Example configuration causing this error:

variable "vcn_dns_label" {
  type  = "string"
  default = "vcn"
}

Remove the quotes from type declarations of variables. Example update:

variable "vcn_dns_label" {
  type = string
  default = "vcn"
}

Error: Error in function call (map)

The log shows an error message similar to the following.

Error: Error in function call
on main.tf line 44, in resource "oci_core_subnet" "this":
44:       display_name        = lookup(map("a", "b", "c", "d"), "a", "default")
────────────────
while calling map(vals...)
Call to function "map" failed: the "map" function was deprecated in Terraform v0.12 and 
is no longer available; use tomap({ ... }) syntax to write a literal map.

The configuration doesn't meet requirements for the specified Terraform version. Version 1.0.x and later don't use this syntax for maps. Example configuration causing this error:

resource "oci_core_subnet" "this" {
   ...
   ...
   vcn_id = lookup(map("a", 1, "b", 2), "a", "default")
   ...
   ...
}

Correct the syntax for the map to use tomap(). Example update:

resource "oci_core_subnet" "this" {
   ...
   ...
   vcn_id = lookup(tomap({"a" = 1, "b" = 2}), "a", "default")
   ...
   ...
}

Error: Error in function call (list)

The log shows an error message similar to the following.

Error: Error in function call
on main.tf line 35, in resource "oci_core_subnet" "this"
  35:       count               = length(list("phx-ad-1"", ""phx-ad-2"))
Call to function "list"" failed: the ""list" function was deprecated in
Terraform v0.12 and is no longer available; use tolist([ ... ]) syntax to
write a literal list.

The configuration doesn't meet requirements for the specified Terraform version. Version 1.0.x and later don't use this syntax for lists. Example configuration causing this error:

resource "oci_core_subnet" "this" {
   count = length(list("phx-ad-1", "phx-ad-2"))
   ...
   ...
}

Correct the syntax for the list to use tolist(). Example update:

resource "oci_core_subnet" "this" {
   count = length(tolist(["phx-ad-1", "phx-ad-2"]))
   ...
   ...
}

Error: The ["*"] form of ignore_changes wildcard is deprecated

The log shows an error message similar to the following.

Getting providers from registry and/or custom terraform providers
resource "oci_core_subnet" "this"
  44:         ignore_changes = ["*"]
The ["*"] form of ignore_changes wildcard is was deprecated and is now
invalid. Use "ignore_changes = all" to ignore changes to all attributes.

The configuration doesn't meet requirements for the specified Terraform version. Version 1.0.x and later don't use this syntax for ignore_changes wildcards. Example configuration causing this error:

resource "oci_core_subnet" "this" {
    ...
    ...
    lifecycle {
      ignore_changes = ["*"]
    }
}

Use ignore_changes = all instead. Example update:

resource "oci_core_subnet" "this" {
    ...
    ...
    lifecycle {
      ignore_changes = all
    }
}

Issue: Deprecated HCL Syntax

The log indicates existence of deprecated HCL syntax.

The configuration doesn't meet requirements for the specified Terraform version.

Update the configuration to omit deprecated HCL syntax.

Oracle Cloud Infrastructure Documentation

Upgrading a Stack to a Later Terraform Version

Before You Begin

Automatic Upgrade

Upgrade Paths

Task 1: Confirm Up-to-Date Infrastructure

Task 2: Download the Configuration and State

Task 3: Upgrade the Configuration

Task 4: Upgrade the Stack

Task 5: Import the State

Task 6: Check for Issues

Troubleshooting Logs During an Upgrade

Error: Failed to install providers

Error: Invalid quoted type constraints

Error: Error in function call (map)

Error: Error in function call (list)

Error: The ["*"] form of ignore_changes wildcard is deprecated

Issue: Deprecated HCL Syntax