aws_workspace: Timeout reached on creation

[ghost]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or other comments that do not add relevant new information or questions, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Terraform Version

Terraform v0.12.24

Affected Resource(s)

• aws_workspace

Terraform Configuration Files

# Copy-paste your Terraform configurations here - for large Terraform configs,
# please use a service like Dropbox and share a link to the ZIP file. For
# security, you can also encrypt the files using our GPG public key: https://keybase.io/hashicorp

resource "aws_workspaces_workspace" "workspace_racker" {
directory_id = module.directory.directory_id
bundle_id = data.aws_workspaces_bundle.standard_windows_10.id
user_name = "racker-test"

workspace_properties {
compute_type_name = data.aws_workspaces_bundle.standard_windows_10.compute_type[0].name
user_volume_size_gib = data.aws_workspaces_bundle.standard_windows_10.user_storage[0].capacity
root_volume_size_gib = data.aws_workspaces_bundle.standard_windows_10.root_storage[0].capacity
running_mode = "AUTO_STOP"
running_mode_auto_stop_timeout_in_minutes = 60
}

tags = {
Environment = "Development"
}

}

Debug Output

aws_workspaces_workspace.workspace_racker: Still creating... [29m1s elapsed]
aws_workspaces_workspace.workspace_racker: Still creating... [29m11s elapsed]
aws_workspaces_workspace.workspace_racker: Still creating... [29m21s elapsed]
aws_workspaces_workspace.workspace_racker: Still creating... [29m31s elapsed]
aws_workspaces_workspace.workspace_racker: Still creating... [29m41s elapsed]
aws_workspaces_workspace.workspace_racker: Still creating... [29m51s elapsed]
aws_workspaces_workspace.workspace_racker: Still creating... [30m1s elapsed]

Error: workspace "ws-7dq7plqrv" is not available: timeout while waiting for state to become 'AVAILABLE' (last state: 'PENDING', timeout: 30m0s)

on main.tf line 134, in resource "aws_workspaces_workspace" "workspace_racker":
134: resource "aws_workspaces_workspace" "workspace_racker" {

Panic Output

Expected Behavior

Workspace creation should be successful

Actual Behavior

Workspace creation reaches a timeout and the state is not updated successfully. Either the internal timeout should be updated or the nested block timeouts should be allowed, just like RDS

Steps to Reproduce

1. terraform apply

Important Factoids

References

• #0000

[fweep]

This is a critical usability change. In us-east-1, there's about a 50-50 chance that a Workspace create will timeout. It's unusable for automation in its current state. Successful deploys take at least 20 minutes, and often skirt the edge of 30 minutes and just squeeze under. I don't know if other regions are faster, but this simply cannot be used for automated deploys in us-east-1 in its current state.

[bflad]

Since this seems valid that the default timeout is too low. If anyone here would like to adjust this, it can be done by updating WorkspaceAvailableTimeout in aws/internal/service/workspaces/waiter/waiter.go to say, 60 minutes (unless the AWS service documentation provides better guidance on how long this can take).

[fweep]

It would be better if it could be configured. Really, all resources should support adjustable timeouts, the way e.g. aws_db_instance does. Many resources in AWS take over 20 minutes to create at times, particularly Windows resources. I've been struggling with timeouts on aws_db_instance, aws_workspaces_workspace, aws_directory_service_directory, and others. AWS Microsoft Managed AD is particularly long-running, and multi-AZ SQL Server instances can take upwards of an hour to replace, especially if final snapshots are enabled. I was surprised to learn that most resources don't support modifying timeouts.

[sulabhk]

+1, I second @fweep in having a configurable option wrt timeouts not just for WS, but for other services like RDS, Aurora etc that take longer times to be created/updated.

[aleliaert]

+1, am helping a Windows IT team move to the cloud with an infrastructure-as-code mindset. Don't want to retreat and use CloudFormation. Help appreciated!

[Tensho]

Hi there, I've recently added timeout attributes to aws_workspaces_workspace resource. Please could you test them for your configuration? 🙇

[bflad]

Given the resource now supports configurable timeouts and since there has been no further activity here, I'm going to close this issue.

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. Thanks!