Deploy error reporting is not showing the reason of failure when using nested stacks.

[xoadrian]

main/root stack is not properly reporting errors from nested stacks, there is no way to know why it actually failed because the error is generalized that some resources could not be created, but why, there is no reason. The only chance to see the error is to constantly monitor the deployment on AWS CloudFormation Console, and once it starts to roll back and delete nested stacks you can quickly see why the nested stack failed in events tab.

E.g.:

creating CloudFormation changeset...
 0/3 | 2:07:56 PM | CREATE_IN_PROGRESS   | AWS::CDK::Metadata         | CDKMetadata
 0/3 | 2:07:56 PM | CREATE_IN_PROGRESS   | AWS::CloudFormation::Stack | appsync-bizon.NestedStack/appsync-bizon.NestedStackResource (appsyncbizonNestedStackappsyncbizonNestedStackResource814A53CF)
 0/3 | 2:07:57 PM | CREATE_IN_PROGRESS   | AWS::CloudFormation::Stack | appsync-bizon.NestedStack/appsync-bizon.NestedStackResource (appsyncbizonNestedStackappsyncbizonNestedStackResource814A53CF) Resource creation Initiated
 0/3 | 2:07:58 PM | CREATE_IN_PROGRESS   | AWS::CDK::Metadata         | CDKMetadata Resource creation Initiated
 1/3 | 2:07:58 PM | CREATE_COMPLETE      | AWS::CDK::Metadata         | CDKMetadata
 2/3 | 2:08:19 PM | CREATE_FAILED        | AWS::CloudFormation::Stack | appsync-bizon.NestedStack/appsync-bizon.NestedStackResource (appsyncbizonNestedStackappsyncbizonNestedStackResource814A53CF) Embedded stack arn:aws:cloudformation:eu-west-2:452364846299:stack/ProductName-bizon-appsyncbizonNestedStackappsyncbizonNestedStackResource814A53CF-4NAE12ZLOQGT/a92a4b40-3c46-11ea-9fec-06452e71541c was not successfully created: The following resource(s) failed to create: [dynamodbstackbizonNestedStackdynamodbstackbizonNestedStackResource0AACEAF0, productnameresolversNestedStackproductnameresolversNestedStackResource42DF9B6A].

as you can see there is no reason why it failed, just saying that some resources failed to create. While actually the error on the nested stack events tab is:

E.g.:

Service role arn not specified (Service: AWSAppSync; Status Code: 400; Error Code: BadRequestException; Request ID: 5a9b00be-f21c-4cea-97ad-3d0e8552c5e5)

or

Only one resolver is allowed per field. (Service: AWSAppSync; Status Code: 400; Error Code: BadRequestException; Request ID: 39813c06-7d28-4d72-be32-f05785f4338e)
  QueryListEvents, WikiArticleLastRevision

or

Property validation failure: [Value for property {/Name} does not match pattern {[_A-Za-z][_0-9A-Za-z]*}]

So there is no way to know about those errors because on roll back those stacks are removed, and developer will never know why it failed.

Reproduction Steps

Just add a nested stack and try to simulate the above error examples:

1. Create AppSync nested stack but don't create service roll for it, and try to create another resource which should rely on that service role.
2. Name a resource in a nested stack improperly with - or .
3. Create 2 same resolver in one nested stack.
   etc ... basically any error that will be thrown inside nested stack

Error Log

already specified above

Environment

• CLI Version : 1.21.0
• Framework Version:
• OS : Windows
• Language : Typescript

Other

I found this issue opened as a feature #4489 but I don't think it is a feature, rather it is a bug, since the feature of having Nested Stacks exists, this means that error reporting for nested stack to be chained in the parent wasn't.

---

This is 🐛 Bug Report

[VictorMorenoJimenez]

Hi, any updates on this? I'm working with nested stacks and it is pretty annoying to debug errors. Thanks!

[nlaffey-icario]

I'm also trying to learn CDK right now and wondering how I'm supposed to debug anything when all I'm getting is something like "The following resource(s) failed to create: [Serverapp, EbsStackawselasticbeanstalkec2roleD9735CA4]."

[ghost]

News on this? Thanks!

[cosmin5mins]

Still an issue for us

[karlskellenPFG]

wack

[Zackhardtoname]

Just to highlight the OP's solution

The only chance to see the error is to constantly monitor the deployment on AWS CloudFormation Console

It actually worked for and unblocked me

[RcSepp]

I faced the same issue, but found the failed nested stack in the CloudFormation CLI under stacks and when toggling the dropdown from "Active" to "Deleted". The stack events for that one contained the failure reason.

[songzh-AZ]

same issue still, any dev willing to look for this?

[fredtcaroli]

Can we get an update on this? Super annoying to debug

[DavidDryja]

Update please? It would help to see an error when debugging.

[Aeolun]

You can actually get the error on the CloudFormation console by going to the parent stack, finding the event for the child stack that failed, copying the name of the stack. Going to the sidebar and selecting 'Deleted Stacks', then searching for the child stack name.

Then look through the events of the deleted child stack to find the actual issue.

This is very far from user friendly.

[cscetbon]

@TheRealAmazonKendra It seems the closed PR didn't fix the issue. Any update on this ? it's been 2 years ...

[LeoStage]

similar issues here, I'm receiving an error when try to deploy and getting this message
❌ Deployment failed: Error: The stack named xxx failed to deploy: UPDATE_ROLLBACK_COMPLETE: Received response status [FAILED] from custom resource. Message returned: See the details in CloudWatch Log Stream: 2023/09/25/[$LATEST]1865cd90f7eb4da6aac8239dad6573d4 (RequestId: b550d905-e34b-41f9-a9b6-0c87986425f0)

When I went to the logs there is no information about the error

[Amplifiyer]

Change #27318 is released with aws-cdk 2.100.0. Can you test with this version and see if this issue is resolved?

[StevenGBrown]

Change #27318 is released with aws-cdk 2.100.0. Can you test with this version and see if this issue is resolved?

Worked for me with aws-cdk 2.123.0.

[jpbndl]

If it is CREATE_FAILED and it happened to rollback, please filter the CloudFormation list and select "Deleted". The actual reason will be in this stack which has been rolled back or been deleted.