Quikstart datadog_integration_api_call_v2.yaml parameters were updated but the custom resource itself does not support update

[vilikin]

Expected Behavior

I would have expected datadog_integration_api_call_v2.yaml to not require updates, but now our stack is requiring this resource to be updated because the parameters were changed in this commit.

Additionally, it was made in the same commit to not allow updates at all, but to fail on them instead of being a silent no-op like previously. So now our pipelines are stuck as the stack would need to be updated, but it does not support updates.

I would expect such bigger changes to be applied in a new version of the template. And I fail to understand the rationale behind the change: requiring update but not allowing update at the same time? Could you elaborate on the rationale behind the change and how customer's are expected to handle this situation? @sabiurr

Actual Behavior

Stack requires update because of changed parameters, but does not support update.

Steps to Reproduce the Problem

1. Deploy previous version of a stack based on datadog_integration_api_call_v2.yaml
2. Try to update the stack (pulling the latest version of the file)
3. See it fail because it does not support update

Specifications

• Datadog CloudFormation template version: datadog_integration_api_call_v2.yaml

Stacktrace

Update not supported, no operation performed.

[andyshinn]

Noticing this as well. I raised https://github.com/DataDog/cloudformation-template/issues/86 related to this as an ask to get stuff versioned in some way that we can pull templates that don't get updated from under our noses.

[JustinLiang90]

Noticed as well - Here is the PR introducing it. https://github.com/DataDog/cloudformation-template/commit/98896d2002c2f67c8ff517e6214c9a59fefaf08c

[PPan1215]

This is occurring to me too and this is blocking my other parts of CDK stack from updating.

[danny-fairly]

@sabiurr this is blocking our CDK stack from updating as well. Could you advise? Looks to be related to https://github.com/DataDog/cloudformation-template/issues/107 as well

[ribhavjain]

This is occurring for us too and we've seen some behavior occasionally deleting our stack entirely (and therefore the integration) after failed updates.

[sabiurr]

This has been reverted https://github.com/DataDog/cloudformation-template/pull/111

Can you try again and see if it works?

[danny-fairly]

Thanks @sabiurr. It's working again. Note, the stack did have to be deleted and recreated, which means any configurations to the AWS integration via the Datadog UI need to be re-applied.