Add a timeouts block to aws_cloudformation_stack resource

[micahhausler]

aws_cloudformation_stack resources currently have the timeout of 30 minutes, but if you have an auto-scale group manged by CF with an update policy with a PauseTime of 10minutes, you'll go over the Terraform timeout if you have 3 nodes.

...
    "UpdatePolicy": {
       "AutoScalingRollingUpdate": {
           "MaxBatchSize": 1,
            "MinInstancesInService": 2,
            "PauseTime": "PT10M",
            "WaitOnResourceSignals": "false"
        }
    }
...

This is the message you'll get.

* aws_cloudformation_stack.etcd_stack: timeout while waiting for state to become 'CREATE_COMPLETE, UPDATE_COMPLETE, UPDATE_ROLLBACK_COMPLETE, UPDATE_ROLLBACK_FAILED' (last state: 'UPDATE_IN_PROGRESS', timeout: 30m0s)

It'd be great to have aws_cloudformation_stacks with configurable timeouts

timeouts {
    update = "45m"
}

Terraform Version

Terraform v0.9.8

[radeksimko]

Hi @micahhausler
we already allow you to specify custom timeout via timeout_in_minutes. This is API-level timeout passed to the CF API and we also use it when gathering the stack events and looking for failures, we just add 5 more minutes on top of that custom timeout internally as there may be some operations before/after stack creation which we may need to catch (e.g. if CF reaches the timeout and we have to catch that error message caused by that).

I agree that timeouts stanza would be more conventional, now that we have it supported in the schema, so I'm labelling this as enhancement, but the functionality is there, IMO.

Let me know if I missed something.

[micahhausler]

Sounds good to me. Thanks!

[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!